dataset
stringclasses 6
values | model
stringclasses 13
values | seed
int64 42
46
| n_topics
int64 10
50
| topic_descriptions
sequencelengths 1
50
| runtime_s
float64 0.93
52.3k
| encoder
stringclasses 5
values | diversity
float64 0.09
1
| c_npmi
float64 -0.38
0.21
| wec_ex
float64 0.11
0.49
| wec_in
float64 0.07
0.94
|
---|---|---|---|---|---|---|---|---|---|---|
ArXiv ML Papers | ZeroShotTM | 42 | 20 | [
[
"deep",
"based",
"to",
"data",
"and",
"segmentation",
"network",
"time",
"for",
"traffic"
],
[
"attack",
"box",
"attacks",
"robustness",
"training",
"robust",
"adversarial",
"perturbations",
"defense",
"models"
],
[
"order",
"stochastic",
"gradient",
"algorithm",
"bandit",
"epsilon",
"optimization",
"regret",
"convex",
"for"
],
[
"graphs",
"nodes",
"node",
"graph",
"clustering",
"link",
"structure",
"network",
"methods",
"embedding"
],
[
"scores",
"gender",
"word",
"text",
"language",
"semantic",
"attention",
"segmentation",
"net",
"generation"
],
[
"graph",
"video",
"feature",
"attention",
"net",
"temporal",
"node",
"convolution",
"convolutional",
"representation"
],
[
"design",
"necessarily",
"broad",
"98",
"possible",
"service",
"expect",
"provided",
"accuracies",
"imbalance"
],
[
"regression",
"kernel",
"linear",
"estimation",
"approximation",
"computational",
"estimator",
"rank",
"tensor",
"variance"
],
[
"from",
"patients",
"were",
"was",
"vehicle",
"developed",
"and",
"of",
"ai",
"software"
],
[
"our",
"data",
"unsupervised",
"gan",
"model",
"networks",
"loss",
"distribution",
"we",
"which"
],
[
"to",
"for",
"model",
"models",
"we",
"that",
"and",
"in",
"the",
"of"
],
[
"tasks",
"training",
"task",
"visual",
"on",
"image",
"language",
"pre",
"speech",
"downstream"
],
[
"latent",
"bayesian",
"output",
"robust",
"variational",
"inference",
"posterior",
"gaussian",
"approximate",
"distribution"
],
[
"forecasting",
"service",
"pooling",
"forecasts",
"decades",
"rigorous",
"focuses",
"stream",
"contains",
"besides"
],
[
"model",
"in",
"source",
"on",
"models",
"to",
"from",
"task",
"that",
"transfer"
],
[
"learning",
"communication",
"that",
"meta",
"this",
"the",
"agents",
"agent",
"is",
"in"
],
[
"the",
"is",
"of",
"queries",
"in",
"algorithms",
"that",
"divergence",
"for",
"data"
],
[
"data",
"time",
"of",
"the",
"are",
"is",
"method",
"in",
"series",
"sensor"
],
[
"actions",
"learning",
"policy",
"control",
"planning",
"robot",
"rl",
"reinforcement",
"environment",
"environments"
],
[
"the",
"is",
"in",
"data",
"by",
"of",
"temporal",
"and",
"to",
"as"
]
] | 28.004794 | intfloat/e5-large-v2 | 0.73 | -0.033195 | 0.140436 | 0.730432 |
ArXiv ML Papers | ZeroShotTM | 42 | 30 | [
[
"treatment",
"for",
"is",
"in",
"data",
"of",
"the",
"distribution",
"noise",
"dimensional"
],
[
"is",
"of",
"that",
"the",
"we",
"queries",
"in",
"quantum",
"wise",
"for"
],
[
"and",
"research",
"patients",
"of",
"health",
"machine",
"forest",
"disease",
"the",
"screening"
],
[
"the",
"for",
"of",
"prediction",
"data",
"is",
"network",
"methods",
"as",
"by"
],
[
"with",
"neural",
"to",
"we",
"networks",
"image",
"models",
"cnns",
"images",
"our"
],
[
"graphical",
"inference",
"variational",
"sparse",
"bayesian",
"distributions",
"tensor",
"computational",
"conditional",
"large"
],
[
"attack",
"robustness",
"defense",
"against",
"adversarial",
"attacks",
"examples",
"perturbations",
"robust",
"box"
],
[
"speech",
"tasks",
"language",
"visual",
"task",
"reasoning",
"transformer",
"pre",
"downstream",
"recognition"
],
[
"learning",
"interaction",
"paper",
"feature",
"interactions",
"methods",
"data",
"datasets",
"nodes",
"domain"
],
[
"example",
"key",
"concept",
"path",
"numbers",
"called",
"rather",
"evolution",
"recall",
"represent"
],
[
"existing",
"graphs",
"attention",
"gnns",
"graph",
"self",
"sequences",
"nodes",
"benchmarks",
"representation"
],
[
"policies",
"actions",
"reward",
"learning",
"policy",
"reinforcement",
"rl",
"action",
"environment",
"control"
],
[
"assessment",
"segmentation",
"imaging",
"medical",
"computer",
"clinical",
"database",
"documents",
"net",
"utilized"
],
[
"to",
"are",
"models",
"and",
"machine",
"their",
"human",
"research",
"them",
"from"
],
[
"94",
"day",
"discuss",
"publicly",
"curve",
"lead",
"incorporate",
"frameworks",
"profile",
"third"
],
[
"this",
"proposed",
"data",
"matrix",
"algorithm",
"which",
"clustering",
"label",
"the",
"is"
],
[
"we",
"that",
"is",
"agents",
"of",
"for",
"sample",
"bound",
"algorithm",
"the"
],
[
"graph",
"attention",
"node",
"gnns",
"graphs",
"structure",
"representation",
"link",
"source",
"temporal"
],
[
"neural",
"training",
"model",
"distributions",
"have",
"loss",
"data",
"distribution",
"networks",
"network"
],
[
"to",
"knowledge",
"teacher",
"learning",
"on",
"domain",
"items",
"graphs",
"task",
"multiple"
],
[
"and",
"gan",
"data",
"of",
"we",
"networks",
"network",
"neural",
"to",
"in"
],
[
"estimation",
"gaussian",
"estimator",
"approximation",
"bayesian",
"approximations",
"posterior",
"bound",
"variance",
"variational"
],
[
"output",
"means",
"factor",
"expression",
"small",
"dependent",
"relu",
"values",
"adversary",
"quantum"
],
[
"and",
"in",
"human",
"machine",
"challenges",
"are",
"ai",
"research",
"software",
"of"
],
[
"neural",
"without",
"inputs",
"fully",
"pipeline",
"classifiers",
"scores",
"architectures",
"detailed",
"higher"
],
[
"in",
"of",
"the",
"and",
"is",
"cnn",
"to",
"with",
"proposed",
"based"
],
[
"adversarial",
"training",
"model",
"to",
"domain",
"source",
"target",
"networks",
"neural",
"self"
],
[
"gradient",
"convex",
"bandit",
"stochastic",
"optimization",
"convergence",
"regret",
"epsilon",
"descent",
"order"
],
[
"series",
"time",
"the",
"of",
"data",
"and",
"for",
"method",
"by",
"based"
],
[
"learning",
"rl",
"this",
"that",
"the",
"is",
"policy",
"algorithms",
"in",
"which"
]
] | 28.611131 | intfloat/e5-large-v2 | 0.66 | -0.049608 | 0.135955 | 0.711752 |
ArXiv ML Papers | ZeroShotTM | 42 | 40 | [
[
"data",
"gan",
"is",
"to",
"network",
"distribution",
"networks",
"the",
"neural",
"federated"
],
[
"stochastic",
"for",
"regret",
"gradient",
"convex",
"bandit",
"convergence",
"optimization",
"gap",
"epsilon"
],
[
"segmentation",
"processing",
"net",
"evaluation",
"layer",
"medical",
"resolution",
"architectures",
"computer",
"map"
],
[
"the",
"and",
"of",
"is",
"in",
"for",
"by",
"with",
"an",
"model"
],
[
"reasoning",
"tasks",
"task",
"model",
"on",
"visual",
"language",
"text",
"recognition",
"word"
],
[
"ai",
"variables",
"imputation",
"and",
"the",
"intelligence",
"forest",
"set",
"cancer",
"of"
],
[
"whose",
"relying",
"combination",
"xgboost",
"key",
"invariance",
"distributed",
"computationally",
"defining",
"being"
],
[
"gnns",
"net",
"convolutional",
"nas",
"including",
"networks",
"search",
"topological",
"autoencoder",
"recurrent"
],
[
"learning",
"which",
"graph",
"proposed",
"manifold",
"nodes",
"multi",
"algorithms",
"graphs",
"label"
],
[
"normalization",
"demonstrate",
"tensor",
"imitation",
"recent",
"hierarchical",
"produce",
"magnitude",
"semi",
"ensemble"
],
[
"policy",
"control",
"environment",
"reinforcement",
"rl",
"tasks",
"action",
"actions",
"imitation",
"reward"
],
[
"laplacian",
"correct",
"volume",
"thus",
"rules",
"corresponding",
"approximately",
"exact",
"degrade",
"sufficiently"
],
[
"the",
"neural",
"network",
"is",
"with",
"to",
"of",
"training",
"and",
"on"
],
[
"policy",
"learning",
"agent",
"reinforcement",
"rl",
"agents",
"reward",
"goal",
"environment",
"that"
],
[
"to",
"data",
"and",
"detection",
"of",
"in",
"with",
"object",
"the",
"time"
],
[
"mode",
"phenomena",
"94",
"those",
"eeg",
"insufficient",
"associated",
"stream",
"considered",
"2015"
],
[
"robot",
"scale",
"learning",
"task",
"from",
"domain",
"teacher",
"tasks",
"transfer",
"knowledge"
],
[
"the",
"algorithms",
"is",
"in",
"matrix",
"for",
"which",
"of",
"rank",
"that"
],
[
"tasks",
"to",
"pre",
"trained",
"our",
"on",
"models",
"that",
"brain",
"an"
],
[
"models",
"robustness",
"adversarial",
"dnns",
"attacks",
"training",
"attack",
"perturbation",
"we",
"robust"
],
[
"modern",
"patch",
"extreme",
"weather",
"period",
"run",
"employed",
"demonstrates",
"explanation",
"svm"
],
[
"hardware",
"neural",
"resolution",
"net",
"segmentation",
"end",
"recognition",
"convolutional",
"attention",
"network"
],
[
"algorithm",
"estimator",
"regression",
"convex",
"estimation",
"bound",
"rank",
"convergence",
"under",
"number"
],
[
"embeddings",
"speech",
"text",
"word",
"corpora",
"language",
"languages",
"audio",
"speaker",
"recognition"
],
[
"user",
"models",
"users",
"bias",
"face",
"media",
"not",
"recognition",
"text",
"systems"
],
[
"machine",
"social",
"development",
"ml",
"research",
"researchers",
"and",
"this",
"questions",
"challenges"
],
[
"matrix",
"algorithms",
"of",
"problem",
"in",
"is",
"clustering",
"graph",
"methods",
"data"
],
[
"regression",
"inference",
"kernel",
"gaussian",
"linear",
"regularization",
"distributions",
"approach",
"posterior",
"estimation"
],
[
"that",
"is",
"of",
"with",
"the",
"strategies",
"we",
"learning",
"regret",
"agent"
],
[
"attack",
"adversarial",
"training",
"examples",
"attacks",
"against",
"robustness",
"robust",
"perturbations",
"box"
],
[
"weights",
"training",
"can",
"loss",
"datasets",
"to",
"networks",
"learning",
"propose",
"we"
],
[
"to",
"model",
"based",
"detection",
"models",
"in",
"and",
"data",
"of",
"monitoring"
],
[
"distribution",
"for",
"it",
"time",
"in",
"the",
"is",
"of",
"test",
"method"
],
[
"graph",
"graphs",
"learning",
"supervised",
"methods",
"knowledge",
"domain",
"from",
"framework",
"embedding"
],
[
"graphs",
"domain",
"embeddings",
"feature",
"view",
"structure",
"graph",
"node",
"nodes",
"embedding"
],
[
"of",
"proposed",
"the",
"estimation",
"is",
"methods",
"method",
"classification",
"data",
"in"
],
[
"linear",
"algorithm",
"gradient",
"function",
"approximation",
"estimator",
"stochastic",
"convergence",
"optimal",
"functions"
],
[
"on",
"representations",
"representation",
"transformer",
"tasks",
"universal",
"image",
"sequence",
"attention",
"model"
],
[
"data",
"learning",
"deep",
"models",
"model",
"we",
"based",
"traffic",
"to",
"time"
],
[
"variational",
"inference",
"distribution",
"posterior",
"bayesian",
"variance",
"markov",
"latent",
"structured",
"approximation"
]
] | 32.388485 | intfloat/e5-large-v2 | 0.6075 | -0.049953 | 0.134342 | 0.728812 |
ArXiv ML Papers | ZeroShotTM | 42 | 50 | [
[
"considered",
"evolution",
"design",
"example",
"correct",
"20",
"identify",
"invariance",
"finally",
"pooling"
],
[
"convex",
"epsilon",
"bandit",
"regret",
"minimax",
"optimization",
"online",
"gradient",
"known",
"an"
],
[
"as",
"of",
"images",
"deep",
"for",
"the",
"segmentation",
"time",
"is",
"to"
],
[
"training",
"networks",
"in",
"to",
"neural",
"have",
"with",
"data",
"our",
"network"
],
[
"given",
"the",
"for",
"of",
"test",
"is",
"distribution",
"divergence",
"in",
"matrix"
],
[
"estimation",
"functions",
"bayesian",
"uncertainty",
"sparsity",
"kernel",
"inference",
"optimization",
"dimensional",
"approximate"
],
[
"training",
"supervised",
"domain",
"representations",
"image",
"datasets",
"adversarial",
"self",
"unsupervised",
"tasks"
],
[
"in",
"adversarial",
"user",
"learning",
"that",
"robustness",
"strategies",
"of",
"to",
"the"
],
[
"mcmc",
"variational",
"robust",
"nearest",
"coding",
"mean",
"field",
"random",
"priors",
"version"
],
[
"semantic",
"resolution",
"execution",
"previously",
"net",
"events",
"surrogate",
"heuristics",
"f1",
"25"
],
[
"classification",
"data",
"the",
"of",
"time",
"and",
"by",
"for",
"variability",
"detection"
],
[
"classification",
"is",
"an",
"to",
"and",
"in",
"the",
"of",
"model",
"data"
],
[
"against",
"attack",
"attacks",
"adversarial",
"robot",
"robustness",
"training",
"learning",
"model",
"to"
],
[
"graph",
"graphs",
"networks",
"node",
"neural",
"structure",
"gnns",
"over",
"neighborhood",
"search"
],
[
"speech",
"attention",
"representation",
"sequence",
"transformer",
"speaker",
"word",
"text",
"semantic",
"resolution"
],
[
"iteration",
"gaussian",
"sampling",
"linear",
"variational",
"estimator",
"variance",
"posterior",
"dimension",
"approximation"
],
[
"algorithms",
"is",
"that",
"regret",
"algorithm",
"we",
"learning",
"decision",
"of",
"an"
],
[
"interaction",
"research",
"human",
"fairness",
"machine",
"these",
"this",
"ml",
"their",
"decisions"
],
[
"training",
"learning",
"model",
"teacher",
"adversarial",
"we",
"perturbations",
"attack",
"to",
"student"
],
[
"training",
"detection",
"have",
"models",
"model",
"to",
"are",
"on",
"samples",
"adversarial"
],
[
"weights",
"neural",
"methods",
"network",
"function",
"non",
"pruning",
"networks",
"optimization",
"sparse"
],
[
"method",
"of",
"for",
"statistics",
"kernel",
"time",
"is",
"series",
"are",
"the"
],
[
"architecture",
"networks",
"end",
"recurrent",
"net",
"hardware",
"accelerators",
"architectures",
"layers",
"performance"
],
[
"self",
"convolutions",
"graph",
"convolution",
"features",
"temporal",
"attention",
"feature",
"node",
"classes"
],
[
"data",
"image",
"performance",
"based",
"segmentation",
"deep",
"images",
"reconstruction",
"3d",
"imaging"
],
[
"of",
"and",
"to",
"in",
"detector",
"by",
"sensor",
"the",
"with",
"performance"
],
[
"predicted",
"autoencoders",
"patch",
"profile",
"attacker",
"literature",
"just",
"studied",
"theorem",
"heuristics"
],
[
"on",
"attacks",
"are",
"models",
"model",
"language",
"pre",
"trained",
"privacy",
"tasks"
],
[
"for",
"that",
"is",
"and",
"in",
"of",
"the",
"are",
"queries",
"distance"
],
[
"agent",
"reward",
"reinforcement",
"policy",
"rl",
"environment",
"learning",
"agents",
"control",
"actions"
],
[
"nodes",
"graph",
"convolutional",
"rnn",
"spectral",
"architectures",
"neural",
"networks",
"autoencoder",
"end"
],
[
"node",
"graphs",
"graph",
"learning",
"tasks",
"knowledge",
"framework",
"contrastive",
"gnns",
"reinforcement"
],
[
"networks",
"view",
"nas",
"graphs",
"is",
"graph",
"to",
"the",
"ranking",
"architecture"
],
[
"that",
"is",
"of",
"queries",
"in",
"this",
"the",
"by",
"matrix",
"we"
],
[
"costs",
"discovery",
"causal",
"parameters",
"search",
"uncertainty",
"inference",
"activation",
"magnitude",
"introduce"
],
[
"agents",
"policy",
"rl",
"reinforcement",
"agent",
"state",
"reward",
"learning",
"trajectory",
"goal"
],
[
"matrix",
"be",
"rank",
"problem",
"as",
"we",
"show",
"problems",
"sparsity",
"number"
],
[
"of",
"data",
"by",
"the",
"this",
"in",
"that",
"as",
"are",
"is"
],
[
"gradient",
"posterior",
"optimization",
"numerical",
"variational",
"convex",
"variance",
"approximate",
"kernel",
"minimization"
],
[
"detection",
"recognition",
"performance",
"models",
"speech",
"have",
"deep",
"dataset",
"used",
"datasets"
],
[
"word",
"task",
"text",
"visual",
"language",
"transformer",
"downstream",
"representations",
"sequence",
"reasoning"
],
[
"the",
"in",
"of",
"is",
"that",
"this",
"data",
"proposed",
"as",
"by"
],
[
"classifier",
"people",
"day",
"clinical",
"focuses",
"critical",
"overcome",
"measurement",
"profile",
"brain"
],
[
"gnns",
"our",
"network",
"networks",
"learned",
"on",
"is",
"nodes",
"graph",
"based"
],
[
"language",
"bias",
"systems",
"speech",
"face",
"text",
"nlp",
"word",
"languages",
"human"
],
[
"researchers",
"or",
"research",
"and",
"social",
"health",
"ml",
"from",
"challenges",
"prediction"
],
[
"attack",
"adversarial",
"attacks",
"robustness",
"perturbation",
"robust",
"training",
"perturbations",
"against",
"examples"
],
[
"random",
"forest",
"patients",
"from",
"and",
"cancer",
"power",
"were",
"ai",
"derived"
],
[
"of",
"to",
"data",
"an",
"the",
"that",
"and",
"we",
"model",
"based"
],
[
"function",
"stochastic",
"gradient",
"algorithm",
"convex",
"gaussian",
"linear",
"convergence",
"optimization",
"order"
]
] | 31.204861 | intfloat/e5-large-v2 | 0.534 | -0.039115 | 0.139871 | 0.71558 |
BBC News | BERTopic | 42 | 10 | [
[
"of",
"in",
"to",
"and",
"the",
"he",
"for",
"we",
"but",
"his"
],
[
"to",
"said",
"and",
"that",
"the",
"in",
"of",
"for",
"its",
"is"
],
[
"he",
"to",
"and",
"in",
"of",
"the",
"mr",
"said",
"that",
"be"
],
[
"it",
"that",
"the",
"of",
"in",
"to",
"is",
"and",
"for",
"are"
],
[
"on",
"to",
"and",
"for",
"film",
"best",
"of",
"in",
"was",
"the"
],
[
"she",
"her",
"in",
"of",
"and",
"race",
"indoor",
"to",
"the",
"world"
],
[
"the",
"and",
"in",
"but",
"of",
"open",
"his",
"was",
"to",
"for"
]
] | 992.662149 | intfloat/e5-large-v2 | 0.4 | -0.021484 | 0.276068 | 0.563411 |
BBC News | BERTopic | 42 | 20 | [
[
"of",
"in",
"to",
"and",
"the",
"he",
"for",
"we",
"but",
"his"
],
[
"to",
"said",
"and",
"that",
"the",
"in",
"of",
"for",
"its",
"is"
],
[
"he",
"to",
"and",
"in",
"of",
"the",
"mr",
"said",
"that",
"be"
],
[
"it",
"that",
"the",
"of",
"in",
"to",
"is",
"and",
"for",
"are"
],
[
"on",
"to",
"and",
"for",
"film",
"best",
"of",
"in",
"was",
"the"
],
[
"she",
"her",
"in",
"of",
"and",
"race",
"indoor",
"to",
"the",
"world"
],
[
"the",
"and",
"in",
"but",
"of",
"open",
"his",
"was",
"to",
"for"
]
] | 861.165591 | intfloat/e5-large-v2 | 0.4 | -0.021484 | 0.276068 | 0.560458 |
BBC News | BERTopic | 42 | 30 | [
[
"of",
"in",
"to",
"and",
"the",
"he",
"for",
"we",
"but",
"his"
],
[
"to",
"said",
"and",
"that",
"the",
"in",
"of",
"for",
"its",
"is"
],
[
"he",
"to",
"and",
"in",
"of",
"the",
"mr",
"said",
"that",
"be"
],
[
"it",
"that",
"the",
"of",
"in",
"to",
"is",
"and",
"for",
"are"
],
[
"on",
"to",
"and",
"for",
"film",
"best",
"of",
"in",
"was",
"the"
],
[
"she",
"her",
"in",
"of",
"and",
"race",
"indoor",
"to",
"the",
"world"
],
[
"the",
"and",
"in",
"but",
"of",
"open",
"his",
"was",
"to",
"for"
]
] | 1,191.793732 | intfloat/e5-large-v2 | 0.4 | -0.021484 | 0.276068 | 0.559822 |
BBC News | BERTopic | 42 | 40 | [
[
"of",
"in",
"to",
"and",
"the",
"he",
"for",
"we",
"but",
"his"
],
[
"to",
"said",
"and",
"that",
"the",
"in",
"of",
"for",
"its",
"is"
],
[
"he",
"to",
"and",
"in",
"of",
"the",
"mr",
"said",
"that",
"be"
],
[
"it",
"that",
"the",
"of",
"in",
"to",
"is",
"and",
"for",
"are"
],
[
"on",
"to",
"and",
"for",
"film",
"best",
"of",
"in",
"was",
"the"
],
[
"she",
"her",
"in",
"of",
"and",
"race",
"indoor",
"to",
"the",
"world"
],
[
"the",
"and",
"in",
"but",
"of",
"open",
"his",
"was",
"to",
"for"
]
] | 1,138.066424 | intfloat/e5-large-v2 | 0.4 | -0.021484 | 0.276068 | 0.568855 |
BBC News | BERTopic | 42 | 50 | [
[
"of",
"in",
"to",
"and",
"the",
"he",
"for",
"we",
"but",
"his"
],
[
"to",
"said",
"and",
"that",
"the",
"in",
"of",
"for",
"its",
"is"
],
[
"he",
"to",
"and",
"in",
"of",
"the",
"mr",
"said",
"that",
"be"
],
[
"it",
"that",
"the",
"of",
"in",
"to",
"is",
"and",
"for",
"are"
],
[
"on",
"to",
"and",
"for",
"film",
"best",
"of",
"in",
"was",
"the"
],
[
"she",
"her",
"in",
"of",
"and",
"race",
"indoor",
"to",
"the",
"world"
],
[
"the",
"and",
"in",
"but",
"of",
"open",
"his",
"was",
"to",
"for"
]
] | 1,042.900464 | intfloat/e5-large-v2 | 0.4 | -0.021484 | 0.276068 | 0.569021 |
BBC News | NMF | 42 | 10 | [
[
"at",
"by",
"the",
"first",
"which",
"it",
"on",
"to",
"time",
"from"
],
[
"not",
"we",
"to",
"that",
"and",
"is",
"with",
"have",
"of",
"it"
],
[
"be",
"are",
"they",
"by",
"the",
"that",
"to",
"said",
"on",
"have"
],
[
"and",
"his",
"to",
"mr",
"was",
"he",
"said",
"for",
"would",
"party"
],
[
"in",
"was",
"to",
"and",
"with",
"at",
"his",
"but",
"for",
"the"
],
[
"its",
"that",
"was",
"said",
"were",
"of",
"the",
"by",
"in",
"has"
],
[
"with",
"and",
"for",
"of",
"more",
"are",
"is",
"as",
"they",
"on"
],
[
"that",
"is",
"it",
"to",
"music",
"what",
"in",
"you",
"not",
"if"
],
[
"minimum",
"will",
"for",
"at",
"in",
"be",
"is",
"increase",
"and",
"that"
],
[
"for",
"25",
"song",
"robbie",
"of",
"last",
"years",
"best",
"it",
"and"
]
] | 1.791996 | intfloat/e5-large-v2 | 0.49 | -0.001058 | 0.269654 | 0.570727 |
BBC News | NMF | 42 | 20 | [
[
"which",
"the",
"has",
"as",
"to",
"of",
"by",
"at",
"this",
"first"
],
[
"and",
"that",
"we",
"to",
"of",
"is",
"not",
"with",
"have",
"in"
],
[
"the",
"of",
"to",
"by",
"up",
"out",
"and",
"with",
"take",
"over"
],
[
"him",
"silk",
"his",
"he",
"and",
"the",
"has",
"is",
"was",
"party"
],
[
"her",
"was",
"at",
"uk",
"last",
"the",
"year",
"in",
"world",
"their"
],
[
"than",
"was",
"by",
"were",
"an",
"of",
"one",
"new",
"or",
"as"
],
[
"in",
"into",
"and",
"world",
"as",
"work",
"with",
"also",
"other",
"both"
],
[
"is",
"in",
"to",
"it",
"you",
"that",
"but",
"as",
"has",
"not"
],
[
"film",
"year",
"also",
"for",
"at",
"who",
"is",
"best",
"up",
"won"
],
[
"robbie",
"and",
"years",
"for",
"best",
"song",
"25",
"in",
"last",
"music"
],
[
"are",
"they",
"have",
"their",
"on",
"people",
"you",
"is",
"the",
"can"
],
[
"was",
"his",
"but",
"on",
"the",
"with",
"after",
"had",
"it",
"from"
],
[
"be",
"will",
"on",
"would",
"new",
"have",
"in",
"is",
"could",
"at"
],
[
"and",
"she",
"black",
"what",
"to",
"music",
"that",
"her",
"the",
"in"
],
[
"would",
"it",
"said",
"be",
"was",
"and",
"he",
"for",
"had",
"were"
],
[
"minimum",
"for",
"and",
"of",
"to",
"in",
"will",
"that",
"increase",
"are"
],
[
"to",
"for",
"have",
"that",
"in",
"file",
"many",
"firms",
"this",
"software"
],
[
"he",
"blair",
"on",
"mr",
"labour",
"election",
"that",
"party",
"brown",
"was"
],
[
"are",
"mobile",
"more",
"with",
"on",
"of",
"as",
"people",
"tv",
"and"
],
[
"is",
"us",
"in",
"on",
"has",
"its",
"and",
"said",
"by",
"with"
]
] | 2.791277 | intfloat/e5-large-v2 | 0.45 | 0.001947 | 0.236665 | 0.582601 |
BBC News | NMF | 42 | 30 | [
[
"the",
"of",
"to",
"time",
"first",
"only",
"most",
"by",
"made",
"into"
],
[
"that",
"and",
"with",
"in",
"to",
"is",
"of",
"we",
"not",
"be"
],
[
"to",
"work",
"this",
"up",
"make",
"out",
"take",
"of",
"also",
"000"
],
[
"the",
"he",
"his",
"has",
"at",
"him",
"with",
"who",
"was",
"been"
],
[
"world",
"the",
"in",
"year",
"their",
"european",
"first",
"last",
"her",
"uk"
],
[
"new",
"all",
"most",
"part",
"from",
"some",
"of",
"than",
"one",
"an"
],
[
"and",
"but",
"it",
"in",
"you",
"to",
"if",
"the",
"like",
"can"
],
[
"and",
"best",
"it",
"song",
"that",
"in",
"25",
"years",
"but",
"of"
],
[
"has",
"are",
"is",
"be",
"there",
"not",
"but",
"this",
"in",
"you"
],
[
"best",
"the",
"film",
"at",
"her",
"who",
"she",
"won",
"awards",
"award"
],
[
"are",
"their",
"they",
"or",
"were",
"people",
"have",
"them",
"by",
"with"
],
[
"in",
"but",
"that",
"not",
"have",
"we",
"liverpool",
"would",
"are",
"it"
],
[
"be",
"could",
"and",
"but",
"to",
"not",
"should",
"would",
"new",
"law"
],
[
"and",
"music",
"to",
"what",
"she",
"that",
"black",
"as",
"people",
"it"
],
[
"work",
"the",
"also",
"and",
"other",
"world",
"into",
"two",
"both",
"between"
],
[
"to",
"and",
"in",
"for",
"that",
"is",
"minimum",
"of",
"will",
"be"
],
[
"for",
"file",
"technology",
"in",
"that",
"not",
"software",
"and",
"to",
"this"
],
[
"election",
"mr",
"labour",
"blair",
"the",
"howard",
"brown",
"he",
"minister",
"party"
],
[
"silk",
"this",
"for",
"and",
"in",
"to",
"party",
"he",
"has",
"have"
],
[
"in",
"us",
"is",
"company",
"has",
"it",
"by",
"its",
"with",
"yukos"
],
[
"added",
"from",
"up",
"high",
"an",
"000",
"the",
"also",
"for",
"while"
],
[
"after",
"was",
"who",
"the",
"when",
"by",
"but",
"had",
"she",
"were"
],
[
"his",
"with",
"after",
"the",
"from",
"england",
"but",
"at",
"game",
"to"
],
[
"more",
"technology",
"tv",
"to",
"mobile",
"are",
"people",
"with",
"services",
"digital"
],
[
"the",
"which",
"gadget",
"it",
"apple",
"has",
"be",
"gadgets",
"mobile",
"at"
],
[
"will",
"new",
"is",
"have",
"at",
"be",
"by",
"which",
"this",
"next"
],
[
"show",
"on",
"in",
"day",
"some",
"the",
"friday",
"also",
"from",
"which"
],
[
"year",
"from",
"growth",
"at",
"economy",
"economic",
"in",
"us",
"by",
"sales"
],
[
"it",
"mr",
"were",
"said",
"he",
"by",
"the",
"had",
"uk",
"added"
],
[
"such",
"well",
"as",
"to",
"an",
"with",
"more",
"new",
"one",
"however"
]
] | 3.995558 | intfloat/e5-large-v2 | 0.446667 | 0.012502 | 0.23947 | 0.629575 |
BBC News | NMF | 42 | 40 | [
[
"by",
"time",
"made",
"of",
"only",
"most",
"first",
"to",
"the",
"into"
],
[
"are",
"not",
"be",
"is",
"this",
"in",
"to",
"but",
"of",
"there"
],
[
"also",
"up",
"to",
"take",
"over",
"this",
"out",
"make",
"any",
"000"
],
[
"were",
"when",
"in",
"was",
"had",
"the",
"been",
"by",
"after",
"which"
],
[
"the",
"their",
"in",
"world",
"uk",
"european",
"new",
"first",
"final",
"years"
],
[
"new",
"of",
"one",
"all",
"than",
"part",
"and",
"some",
"most",
"many"
],
[
"world",
"into",
"the",
"work",
"and",
"years",
"from",
"also",
"other",
"both"
],
[
"and",
"to",
"best",
"that",
"song",
"of",
"in",
"years",
"it",
"25"
],
[
"new",
"by",
"this",
"year",
"which",
"will",
"be",
"service",
"is",
"have"
],
[
"award",
"won",
"best",
"who",
"director",
"the",
"film",
"awards",
"with",
"actress"
],
[
"of",
"will",
"and",
"in",
"increase",
"for",
"to",
"be",
"that",
"minimum"
],
[
"have",
"we",
"not",
"but",
"liverpool",
"would",
"if",
"had",
"think",
"that"
],
[
"such",
"time",
"to",
"as",
"an",
"new",
"their",
"one",
"well",
"however"
],
[
"music",
"in",
"black",
"the",
"and",
"what",
"that",
"stone",
"she",
"people"
],
[
"the",
"on",
"and",
"from",
"also",
"an",
"day",
"after",
"some",
"friday"
],
[
"government",
"to",
"tax",
"local",
"public",
"in",
"budget",
"would",
"council",
"000"
],
[
"and",
"that",
"to",
"this",
"about",
"file",
"make",
"not",
"on",
"for"
],
[
"mr",
"he",
"told",
"minister",
"prime",
"blair",
"brown",
"the",
"had",
"by"
],
[
"to",
"silk",
"party",
"in",
"he",
"this",
"and",
"for",
"of",
"mr"
],
[
"but",
"if",
"at",
"the",
"like",
"very",
"all",
"it",
"they",
"and"
],
[
"000",
"up",
"added",
"for",
"also",
"an",
"new",
"year",
"news",
"while"
],
[
"firm",
"with",
"its",
"in",
"by",
"company",
"which",
"yukos",
"it",
"oil"
],
[
"we",
"and",
"to",
"not",
"have",
"with",
"of",
"that",
"be",
"government"
],
[
"people",
"the",
"are",
"by",
"and",
"or",
"were",
"their",
"have",
"they"
],
[
"her",
"olympic",
"year",
"with",
"but",
"first",
"she",
"who",
"after",
"had"
],
[
"mini",
"it",
"software",
"apple",
"to",
"pc",
"with",
"mac",
"computer",
"and"
],
[
"his",
"who",
"the",
"him",
"he",
"with",
"but",
"not",
"told",
"would"
],
[
"in",
"year",
"growth",
"by",
"economy",
"from",
"sales",
"prices",
"market",
"than"
],
[
"he",
"the",
"said",
"it",
"were",
"being",
"had",
"uk",
"added",
"rights"
],
[
"would",
"could",
"be",
"not",
"to",
"should",
"have",
"if",
"but",
"new"
],
[
"are",
"technology",
"is",
"of",
"dvd",
"high",
"be",
"on",
"games",
"in"
],
[
"are",
"to",
"mobile",
"technology",
"data",
"3g",
"but",
"networks",
"on",
"of"
],
[
"have",
"number",
"show",
"the",
"at",
"best",
"and",
"his",
"end",
"theatre"
],
[
"howard",
"the",
"labour",
"party",
"election",
"tory",
"would",
"tories",
"tax",
"in"
],
[
"are",
"your",
"with",
"there",
"can",
"and",
"that",
"game",
"if",
"you"
],
[
"been",
"has",
"have",
"this",
"is",
"last",
"said",
"since",
"the",
"now"
],
[
"which",
"mobile",
"of",
"it",
"and",
"the",
"gadget",
"be",
"gadgets",
"in"
],
[
"china",
"with",
"in",
"dollar",
"budget",
"trade",
"us",
"bush",
"deficit",
"about"
],
[
"but",
"with",
"his",
"england",
"after",
"from",
"game",
"half",
"ireland",
"the"
],
[
"people",
"tv",
"with",
"broadband",
"more",
"digital",
"net",
"services",
"service",
"video"
]
] | 5.714477 | intfloat/e5-large-v2 | 0.4575 | 0.020847 | 0.227713 | 0.648393 |
BBC News | NMF | 42 | 50 | [
[
"of",
"and",
"the",
"most",
"time",
"first",
"in",
"only",
"also",
"into"
],
[
"this",
"to",
"is",
"in",
"and",
"but",
"be",
"of",
"not",
"there"
],
[
"mr",
"added",
"of",
"said",
"were",
"and",
"had",
"uk",
"to",
"is"
],
[
"he",
"told",
"him",
"would",
"an",
"but",
"not",
"very",
"added",
"to"
],
[
"first",
"final",
"last",
"years",
"world",
"in",
"uk",
"european",
"new",
"set"
],
[
"and",
"than",
"one",
"of",
"most",
"the",
"to",
"their",
"all",
"also"
],
[
"into",
"two",
"other",
"also",
"and",
"work",
"between",
"both",
"years",
"over"
],
[
"it",
"song",
"of",
"and",
"best",
"25",
"to",
"in",
"years",
"that"
],
[
"make",
"digital",
"to",
"which",
"new",
"be",
"next",
"will",
"can",
"service"
],
[
"the",
"best",
"film",
"awards",
"who",
"award",
"won",
"director",
"actress",
"actor"
],
[
"of",
"to",
"for",
"will",
"minimum",
"increase",
"in",
"and",
"that",
"be"
],
[
"we",
"but",
"liverpool",
"think",
"would",
"had",
"club",
"steven",
"not",
"if"
],
[
"but",
"it",
"the",
"to",
"very",
"at",
"news",
"out",
"because",
"over"
],
[
"and",
"what",
"music",
"to",
"she",
"stone",
"black",
"from",
"or",
"the"
],
[
"that",
"out",
"this",
"not",
"about",
"the",
"to",
"there",
"by",
"file"
],
[
"economic",
"growth",
"year",
"economy",
"from",
"bank",
"rates",
"prices",
"sales",
"market"
],
[
"also",
"from",
"an",
"on",
"the",
"day",
"friday",
"some",
"wednesday",
"media"
],
[
"gadget",
"the",
"be",
"it",
"mobile",
"gadgets",
"which",
"first",
"list",
"or"
],
[
"to",
"party",
"silk",
"and",
"he",
"in",
"this",
"for",
"mr",
"on"
],
[
"their",
"were",
"people",
"they",
"and",
"who",
"them",
"when",
"on",
"those"
],
[
"year",
"while",
"week",
"up",
"new",
"also",
"for",
"an",
"000",
"added"
],
[
"shares",
"which",
"its",
"firm",
"also",
"would",
"london",
"company",
"market",
"deutsche"
],
[
"to",
"of",
"we",
"and",
"that",
"with",
"not",
"be",
"have",
"in"
],
[
"last",
"have",
"been",
"from",
"very",
"may",
"the",
"not",
"this",
"four"
],
[
"had",
"blair",
"mr",
"prime",
"brown",
"his",
"the",
"minister",
"told",
"chancellor"
],
[
"net",
"security",
"to",
"users",
"software",
"or",
"people",
"microsoft",
"attacks",
"site"
],
[
"was",
"had",
"were",
"after",
"been",
"when",
"in",
"but",
"which",
"one"
],
[
"to",
"out",
"this",
"take",
"up",
"make",
"any",
"set",
"which",
"however"
],
[
"the",
"from",
"were",
"by",
"which",
"been",
"000",
"an",
"up",
"number"
],
[
"well",
"as",
"such",
"an",
"time",
"part",
"many",
"however",
"one",
"industry"
],
[
"phones",
"people",
"music",
"more",
"mobile",
"with",
"to",
"phone",
"mobiles",
"the"
],
[
"she",
"who",
"her",
"year",
"the",
"first",
"but",
"after",
"olympic",
"world"
],
[
"number",
"show",
"at",
"the",
"theatre",
"royal",
"top",
"end",
"best",
"expected"
],
[
"party",
"the",
"labour",
"election",
"tory",
"howard",
"would",
"tax",
"tories",
"in"
],
[
"the",
"your",
"if",
"can",
"there",
"what",
"games",
"and",
"game",
"you"
],
[
"now",
"has",
"with",
"been",
"since",
"the",
"said",
"week",
"against",
"internet"
],
[
"it",
"and",
"in",
"of",
"be",
"with",
"have",
"on",
"but",
"children"
],
[
"to",
"the",
"and",
"mobile",
"of",
"3g",
"technology",
"networks",
"data",
"for"
],
[
"england",
"half",
"six",
"after",
"but",
"ireland",
"the",
"game",
"wales",
"their"
],
[
"broadband",
"video",
"people",
"uk",
"content",
"net",
"digital",
"tv",
"services",
"over"
],
[
"of",
"oil",
"in",
"russian",
"yukos",
"its",
"the",
"court",
"gazprom",
"that"
],
[
"be",
"is",
"games",
"the",
"dvd",
"high",
"and",
"technology",
"ray",
"in"
],
[
"mac",
"mini",
"and",
"the",
"it",
"apple",
"pc",
"for",
"computer",
"is"
],
[
"who",
"his",
"him",
"the",
"my",
"after",
"year",
"me",
"old",
"club"
],
[
"can",
"phone",
"companies",
"these",
"be",
"calls",
"people",
"about",
"not",
"is"
],
[
"not",
"law",
"should",
"government",
"be",
"would",
"could",
"lord",
"but",
"if"
],
[
"dollar",
"us",
"china",
"bush",
"trade",
"budget",
"eu",
"deficit",
"about",
"firms"
],
[
"an",
"out",
"the",
"from",
"with",
"over",
"one",
"hit",
"both",
"through"
],
[
"radio",
"in",
"is",
"hop",
"spanish",
"world",
"hip",
"says",
"but",
"rap"
],
[
"where",
"there",
"are",
"be",
"not",
"their",
"which",
"do",
"more",
"they"
]
] | 7.025339 | intfloat/e5-large-v2 | 0.482 | 0.026203 | 0.216141 | 0.681137 |
BBC News | LDA | 42 | 10 | [
[
"and",
"his",
"to",
"the",
"was",
"mr",
"he",
"in",
"of",
"had"
],
[
"ms",
"her",
"the",
"she",
"to",
"in",
"was",
"said",
"air",
"airlines"
],
[
"for",
"to",
"in",
"of",
"that",
"the",
"said",
"and",
"is",
"it"
],
[
"was",
"to",
"the",
"on",
"of",
"in",
"for",
"and",
"is",
"it"
],
[
"to",
"liverpool",
"united",
"the",
"club",
"chelsea",
"smith",
"arsenal",
"he",
"manchester"
],
[
"the",
"in",
"of",
"to",
"said",
"on",
"that",
"be",
"and",
"was"
],
[
"to",
"the",
"and",
"of",
"that",
"it",
"are",
"is",
"in",
"be"
],
[
"it",
"that",
"of",
"be",
"and",
"the",
"in",
"is",
"to",
"for"
],
[
"to",
"the",
"in",
"of",
"and",
"on",
"for",
"said",
"by",
"uk"
],
[
"for",
"and",
"the",
"to",
"but",
"on",
"of",
"in",
"is",
"it"
]
] | 7.619462 | intfloat/e5-large-v2 | 0.33 | -0.026604 | 0.289131 | 0.572182 |
BBC News | LDA | 42 | 20 | [
[
"he",
"in",
"to",
"the",
"was",
"his",
"and",
"it",
"of",
"but"
],
[
"of",
"she",
"was",
"said",
"to",
"her",
"the",
"ms",
"in",
"prince"
],
[
"is",
"it",
"in",
"for",
"the",
"that",
"said",
"of",
"to",
"and"
],
[
"to",
"the",
"and",
"of",
"in",
"for",
"was",
"it",
"on",
"is"
],
[
"and",
"club",
"united",
"we",
"to",
"at",
"the",
"league",
"he",
"has"
],
[
"to",
"in",
"be",
"of",
"was",
"the",
"said",
"and",
"it",
"for"
],
[
"the",
"and",
"that",
"to",
"of",
"said",
"it",
"in",
"is",
"users"
],
[
"that",
"the",
"to",
"and",
"in",
"of",
"is",
"he",
"it",
"on"
],
[
"said",
"and",
"of",
"to",
"in",
"the",
"for",
"by",
"on",
"is"
],
[
"the",
"to",
"in",
"is",
"and",
"for",
"of",
"on",
"at",
"have"
],
[
"the",
"men",
"on",
"cross",
"country",
"arrested",
"in",
"women",
"blue",
"from"
],
[
"the",
"to",
"of",
"and",
"office",
"in",
"linux",
"open",
"its",
"box"
],
[
"at",
"for",
"the",
"best",
"awards",
"theatre",
"actress",
"director",
"actor",
"musical"
],
[
"the",
"of",
"and",
"in",
"will",
"to",
"be",
"on",
"is",
"for"
],
[
"but",
"to",
"and",
"game",
"from",
"was",
"the",
"half",
"on",
"with"
],
[
"and",
"is",
"in",
"of",
"to",
"the",
"that",
"it",
"are",
"for"
],
[
"to",
"the",
"will",
"of",
"and",
"scottish",
"in",
"be",
"is",
"would"
],
[
"dvd",
"the",
"dvds",
"to",
"of",
"definition",
"high",
"be",
"and",
"studios"
],
[
"that",
"of",
"be",
"mr",
"to",
"said",
"in",
"and",
"the",
"was"
],
[
"the",
"of",
"in",
"to",
"and",
"dr",
"mr",
"jewish",
"smith",
"he"
]
] | 16.628456 | intfloat/e5-large-v2 | 0.32 | -0.017112 | 0.259832 | 0.598625 |
BBC News | LDA | 42 | 30 | [
[
"he",
"the",
"to",
"was",
"his",
"had",
"and",
"in",
"of",
"for"
],
[
"the",
"was",
"she",
"in",
"ms",
"her",
"said",
"to",
"of",
"after"
],
[
"of",
"the",
"to",
"in",
"and",
"said",
"that",
"for",
"its",
"is"
],
[
"for",
"in",
"to",
"of",
"the",
"and",
"on",
"was",
"film",
"with"
],
[
"to",
"league",
"chelsea",
"united",
"the",
"club",
"arsenal",
"he",
"manchester",
"on"
],
[
"law",
"of",
"lord",
"in",
"the",
"to",
"said",
"and",
"rights",
"was"
],
[
"to",
"and",
"said",
"the",
"attacks",
"sites",
"of",
"site",
"that",
"in"
],
[
"the",
"and",
"to",
"of",
"he",
"said",
"in",
"on",
"that",
"mr"
],
[
"budget",
"japan",
"in",
"the",
"of",
"bush",
"to",
"and",
"was",
"prince"
],
[
"on",
"and",
"for",
"of",
"is",
"the",
"in",
"to",
"but",
"at"
],
[
"of",
"on",
"from",
"arrested",
"men",
"the",
"women",
"country",
"cross",
"police"
],
[
"and",
"the",
"to",
"ibm",
"software",
"linux",
"programs",
"use",
"of",
"call"
],
[
"the",
"for",
"awards",
"best",
"at",
"award",
"won",
"in",
"theatre",
"actress"
],
[
"at",
"number",
"the",
"to",
"and",
"top",
"chart",
"of",
"in",
"one"
],
[
"to",
"the",
"and",
"was",
"half",
"for",
"game",
"but",
"from",
"after"
],
[
"for",
"the",
"in",
"to",
"people",
"local",
"of",
"and",
"government",
"is"
],
[
"scottish",
"scotland",
"russian",
"executive",
"to",
"will",
"gazprom",
"the",
"and",
"in"
],
[
"the",
"dvds",
"dvd",
"studios",
"film",
"technology",
"to",
"formats",
"of",
"hollywood"
],
[
"of",
"users",
"the",
"to",
"that",
"security",
"and",
"software",
"mail",
"or"
],
[
"gas",
"contract",
"firm",
"oil",
"company",
"production",
"indian",
"contracts",
"the",
"energy"
],
[
"the",
"radio",
"and",
"china",
"in",
"of",
"to",
"on",
"will",
"is"
],
[
"within",
"companies",
"industries",
"television",
"such",
"indicated",
"research",
"led",
"film",
"career"
],
[
"to",
"race",
"world",
"in",
"olympic",
"of",
"the",
"at",
"and",
"indoor"
],
[
"mobile",
"to",
"the",
"phone",
"of",
"and",
"tv",
"people",
"service",
"are"
],
[
"the",
"and",
"to",
"irish",
"000",
"for",
"from",
"u2",
"of",
"drugs"
],
[
"of",
"to",
"the",
"and",
"he",
"has",
"been",
"for",
"in",
"was"
],
[
"it",
"in",
"is",
"for",
"and",
"to",
"that",
"of",
"the",
"as"
],
[
"to",
"the",
"of",
"be",
"for",
"in",
"said",
"and",
"that",
"is"
],
[
"wasps",
"bath",
"ask",
"thompson",
"search",
"leicester",
"computing",
"hardware",
"it",
"web"
],
[
"the",
"to",
"and",
"is",
"of",
"it",
"that",
"we",
"you",
"have"
]
] | 33.410106 | intfloat/e5-large-v2 | 0.446667 | -0.011615 | 0.215939 | 0.655877 |
BBC News | LDA | 42 | 40 | [
[
"party",
"leader",
"he",
"the",
"and",
"mr",
"to",
"his",
"silk",
"as"
],
[
"kim",
"her",
"return",
"the",
"on",
"february",
"in",
"back",
"to",
"make"
],
[
"in",
"the",
"to",
"of",
"for",
"and",
"said",
"that",
"its",
"on"
],
[
"and",
"to",
"film",
"for",
"of",
"on",
"was",
"in",
"at",
"the"
],
[
"the",
"are",
"with",
"brother",
"and",
"big",
"on",
"to",
"life",
"show"
],
[
"to",
"the",
"peer",
"hunting",
"music",
"and",
"that",
"of",
"files",
"hunt"
],
[
"the",
"is",
"to",
"users",
"and",
"software",
"in",
"computer",
"of",
"programs"
],
[
"he",
"of",
"the",
"that",
"and",
"in",
"to",
"mr",
"on",
"is"
],
[
"the",
"broadband",
"japan",
"to",
"japanese",
"by",
"of",
"in",
"and",
"year"
],
[
"to",
"the",
"in",
"and",
"on",
"of",
"for",
"with",
"at",
"was"
],
[
"of",
"the",
"on",
"police",
"men",
"from",
"arrested",
"british",
"to",
"in"
],
[
"him",
"edward",
"for",
"at",
"to",
"prince",
"the",
"queen",
"harry",
"of"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"and",
"to",
"in",
"the",
"super",
"of",
"group",
"said",
"its",
"at"
],
[
"irish",
"the",
"prize",
"is",
"for",
"and",
"000",
"book",
"of",
"novel"
],
[
"and",
"of",
"to",
"people",
"in",
"the",
"that",
"more",
"said",
"music"
],
[
"russia",
"the",
"its",
"state",
"yukos",
"oil",
"firm",
"company",
"gazprom",
"russian"
],
[
"studios",
"dvd",
"gas",
"dvds",
"energy",
"the",
"to",
"india",
"in",
"technology"
],
[
"of",
"the",
"to",
"sport",
"drugs",
"and",
"for",
"that",
"doping",
"collins"
],
[
"security",
"what",
"virus",
"sending",
"anti",
"mail",
"warnings",
"2004",
"itself",
"websites"
],
[
"in",
"of",
"to",
"the",
"on",
"radio",
"digital",
"air",
"and",
"will"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"scottish",
"the",
"in",
"will",
"at",
"to",
"scotland",
"glasgow",
"and",
"list"
],
[
"to",
"mobile",
"the",
"and",
"phone",
"services",
"tv",
"service",
"of",
"will"
],
[
"french",
"album",
"singer",
"band",
"at",
"to",
"the",
"drug",
"music",
"and"
],
[
"in",
"the",
"england",
"to",
"robinson",
"for",
"after",
"has",
"six",
"of"
],
[
"and",
"the",
"in",
"of",
"for",
"it",
"is",
"to",
"that",
"said"
],
[
"to",
"and",
"in",
"is",
"of",
"that",
"the",
"be",
"it",
"for"
],
[
"code",
"or",
"microsoft",
"version",
"online",
"windows",
"system",
"security",
"released",
"pcs"
],
[
"and",
"to",
"it",
"is",
"that",
"the",
"of",
"you",
"not",
"in"
],
[
"the",
"her",
"in",
"she",
"to",
"was",
"has",
"year",
"said",
"but"
],
[
"to",
"search",
"site",
"the",
"users",
"sites",
"web",
"said",
"net",
"google"
],
[
"sporting",
"that",
"the",
"to",
"my",
"manager",
"director",
"and",
"we",
"frank"
],
[
"of",
"budget",
"the",
"to",
"and",
"bush",
"cards",
"be",
"spending",
"said"
],
[
"of",
"said",
"is",
"iraq",
"to",
"and",
"women",
"in",
"the",
"dr"
],
[
"the",
"of",
"in",
"he",
"for",
"to",
"and",
"his",
"it",
"is"
],
[
"in",
"it",
"bill",
"for",
"the",
"children",
"and",
"computing",
"to",
"of"
],
[
"in",
"game",
"it",
"was",
"from",
"the",
"and",
"of",
"to",
"we"
],
[
"the",
"to",
"of",
"in",
"and",
"said",
"he",
"was",
"on",
"that"
],
[
"the",
"of",
"and",
"in",
"as",
"world",
"to",
"said",
"with",
"has"
]
] | 49.44931 | intfloat/e5-large-v2 | 0.415 | -0.03507 | 0.201104 | 0.669696 |
BBC News | LDA | 42 | 50 | [
[
"he",
"of",
"to",
"in",
"the",
"and",
"mr",
"his",
"for",
"said"
],
[
"took",
"number",
"the",
"box",
"last",
"weekend",
"movie",
"office",
"in",
"7m"
],
[
"of",
"said",
"the",
"and",
"to",
"in",
"for",
"that",
"its",
"on"
],
[
"the",
"for",
"on",
"of",
"in",
"to",
"and",
"film",
"was",
"best"
],
[
"manchester",
"ferguson",
"to",
"west",
"united",
"the",
"visit",
"everton",
"city",
"boss"
],
[
"guilty",
"peer",
"network",
"to",
"of",
"the",
"and",
"files",
"sharing",
"that"
],
[
"to",
"india",
"the",
"and",
"of",
"company",
"indian",
"000",
"in",
"it"
],
[
"and",
"it",
"that",
"to",
"of",
"is",
"you",
"the",
"are",
"have"
],
[
"to",
"south",
"countries",
"the",
"korea",
"japan",
"in",
"yen",
"and",
"of"
],
[
"to",
"in",
"the",
"and",
"of",
"on",
"for",
"with",
"was",
"but"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"prince",
"of",
"in",
"harry",
"the",
"and",
"vehicles",
"to",
"said",
"it"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"be",
"and",
"parents",
"group",
"in",
"super",
"to",
"children",
"the",
"of"
],
[
"irish",
"for",
"ireland",
"the",
"penalty",
"gara",
"and",
"try",
"driscoll",
"brian"
],
[
"the",
"for",
"in",
"to",
"of",
"and",
"people",
"more",
"is",
"said"
],
[
"and",
"the",
"he",
"head",
"prop",
"said",
"phil",
"injury",
"in",
"operation"
],
[
"harder",
"system",
"copy",
"piracy",
"copies",
"download",
"dvd",
"said",
"formats",
"new"
],
[
"and",
"to",
"the",
"ban",
"of",
"hunting",
"said",
"that",
"schools",
"hunt"
],
[
"of",
"johnny",
"wasps",
"leicester",
"what",
"the",
"van",
"strange",
"fire",
"murphy"
],
[
"to",
"china",
"the",
"of",
"and",
"in",
"on",
"is",
"chinese",
"be"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"the",
"scottish",
"to",
"will",
"scotland",
"on",
"in",
"be",
"of",
"and"
],
[
"tv",
"and",
"the",
"services",
"phone",
"service",
"of",
"to",
"mobile",
"phones"
],
[
"singer",
"band",
"tour",
"album",
"french",
"to",
"the",
"at",
"and",
"u2"
],
[
"in",
"3g",
"the",
"and",
"for",
"to",
"of",
"drugs",
"has",
"will"
],
[
"the",
"and",
"is",
"of",
"to",
"it",
"in",
"for",
"that",
"said"
],
[
"that",
"for",
"of",
"is",
"in",
"to",
"be",
"it",
"and",
"the"
],
[
"centre",
"robinson",
"coach",
"squad",
"wales",
"england",
"andy",
"six",
"in",
"nations"
],
[
"and",
"of",
"music",
"the",
"it",
"to",
"is",
"that",
"in",
"song"
],
[
"in",
"she",
"was",
"said",
"the",
"to",
"her",
"year",
"it",
"but"
],
[
"search",
"site",
"the",
"users",
"to",
"net",
"sites",
"web",
"of",
"internet"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"of",
"to",
"the",
"in",
"rates",
"and",
"brown",
"spending",
"will",
"economy"
],
[
"and",
"of",
"to",
"in",
"dvd",
"the",
"high",
"sony",
"be",
"is"
],
[
"the",
"his",
"for",
"he",
"to",
"in",
"of",
"it",
"and",
"is"
],
[
"computing",
"of",
"the",
"in",
"hardware",
"to",
"will",
"and",
"it",
"customers"
],
[
"the",
"was",
"of",
"and",
"in",
"it",
"to",
"game",
"we",
"but"
],
[
"and",
"of",
"he",
"the",
"to",
"in",
"was",
"mr",
"his",
"said"
],
[
"security",
"as",
"of",
"in",
"and",
"that",
"to",
"the",
"said",
"are"
],
[
"of",
"world",
"olympic",
"in",
"the",
"to",
"and",
"indoor",
"at",
"britain"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"party",
"and",
"in",
"of",
"to",
"labour",
"the",
"election",
"said",
"he"
],
[
"shot",
"hunting",
"remote",
"session",
"shooting",
"texas",
"control",
"wild",
"camera",
"lets"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"said",
"they",
"to",
"police",
"the",
"and",
"were",
"in",
"have",
"their"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
],
[
"the",
"be",
"on",
"of",
"and",
"mr",
"he",
"said",
"to",
"in"
],
[
"the",
"of",
"to",
"and",
"for",
"in",
"on",
"said",
"be",
"by"
],
[
"points",
"pointed",
"political",
"policy",
"policies",
"police",
"politicians",
"pointing",
"plus",
"pocket"
]
] | 63.239012 | intfloat/e5-large-v2 | 0.362 | -0.068876 | 0.200801 | 0.691685 |
BBC News | Top2Vec | 42 | 10 | [
[
"interested",
"rugby",
"football",
"coaches",
"coaching",
"sports",
"matches",
"defending",
"players",
"opponents"
],
[
"prices",
"business",
"finance",
"investment",
"investors",
"economists",
"businesses",
"finances",
"earnings",
"economics"
],
[
"politics",
"politically",
"suggested",
"regarding",
"politicians",
"political",
"voters",
"asked",
"pensions",
"ideas"
],
[
"internet",
"computers",
"technology",
"gadgets",
"broadband",
"interested",
"websites",
"computing",
"and",
"tech"
],
[
"and",
"from",
"featuring",
"according",
"also",
"oscars",
"awards",
"regarding",
"starring",
"actors"
],
[
"running",
"promising",
"athletes",
"competing",
"olympics",
"athletics",
"qualifying",
"results",
"competitors",
"olympic"
],
[
"sports",
"matches",
"tennis",
"wimbledon",
"rankings",
"results",
"sport",
"competitors",
"competing",
"winning"
]
] | 1,163.078218 | intfloat/e5-large-v2 | 0.885714 | -0.280193 | 0.223161 | 0.906824 |
BBC News | Top2Vec | 42 | 20 | [
[
"interested",
"rugby",
"football",
"coaches",
"coaching",
"sports",
"matches",
"defending",
"players",
"opponents"
],
[
"prices",
"business",
"finance",
"investment",
"investors",
"economists",
"businesses",
"finances",
"earnings",
"economics"
],
[
"politics",
"politically",
"suggested",
"regarding",
"politicians",
"political",
"voters",
"asked",
"pensions",
"ideas"
],
[
"internet",
"computers",
"technology",
"gadgets",
"broadband",
"interested",
"websites",
"computing",
"and",
"tech"
],
[
"and",
"from",
"featuring",
"according",
"also",
"oscars",
"awards",
"regarding",
"starring",
"actors"
],
[
"running",
"promising",
"athletes",
"competing",
"olympics",
"athletics",
"qualifying",
"results",
"competitors",
"olympic"
],
[
"sports",
"matches",
"tennis",
"wimbledon",
"rankings",
"results",
"sport",
"competitors",
"competing",
"winning"
]
] | 1,068.145719 | intfloat/e5-large-v2 | 0.885714 | -0.280193 | 0.223161 | 0.905834 |
BBC News | Top2Vec | 42 | 30 | [
[
"interested",
"rugby",
"football",
"coaches",
"coaching",
"sports",
"matches",
"defending",
"players",
"opponents"
],
[
"prices",
"business",
"finance",
"investment",
"investors",
"economists",
"businesses",
"finances",
"earnings",
"economics"
],
[
"politics",
"politically",
"suggested",
"regarding",
"politicians",
"political",
"voters",
"asked",
"pensions",
"ideas"
],
[
"internet",
"computers",
"technology",
"gadgets",
"broadband",
"interested",
"websites",
"computing",
"and",
"tech"
],
[
"and",
"from",
"featuring",
"according",
"also",
"oscars",
"awards",
"regarding",
"starring",
"actors"
],
[
"running",
"promising",
"athletes",
"competing",
"olympics",
"athletics",
"qualifying",
"results",
"competitors",
"olympic"
],
[
"sports",
"matches",
"tennis",
"wimbledon",
"rankings",
"results",
"sport",
"competitors",
"competing",
"winning"
]
] | 997.269333 | intfloat/e5-large-v2 | 0.885714 | -0.280193 | 0.223161 | 0.912991 |
BBC News | Top2Vec | 42 | 40 | [
[
"interested",
"rugby",
"football",
"coaches",
"coaching",
"sports",
"matches",
"defending",
"players",
"opponents"
],
[
"prices",
"business",
"finance",
"investment",
"investors",
"economists",
"businesses",
"finances",
"earnings",
"economics"
],
[
"politics",
"politically",
"suggested",
"regarding",
"politicians",
"political",
"voters",
"asked",
"pensions",
"ideas"
],
[
"internet",
"computers",
"technology",
"gadgets",
"broadband",
"interested",
"websites",
"computing",
"and",
"tech"
],
[
"and",
"from",
"featuring",
"according",
"also",
"oscars",
"awards",
"regarding",
"starring",
"actors"
],
[
"running",
"promising",
"athletes",
"competing",
"olympics",
"athletics",
"qualifying",
"results",
"competitors",
"olympic"
],
[
"sports",
"matches",
"tennis",
"wimbledon",
"rankings",
"results",
"sport",
"competitors",
"competing",
"winning"
]
] | 944.852174 | intfloat/e5-large-v2 | 0.885714 | -0.280193 | 0.223161 | 0.906171 |
BBC News | Top2Vec | 42 | 50 | [
[
"interested",
"rugby",
"football",
"coaches",
"coaching",
"sports",
"matches",
"defending",
"players",
"opponents"
],
[
"prices",
"business",
"finance",
"investment",
"investors",
"economists",
"businesses",
"finances",
"earnings",
"economics"
],
[
"politics",
"politically",
"suggested",
"regarding",
"politicians",
"political",
"voters",
"asked",
"pensions",
"ideas"
],
[
"internet",
"computers",
"technology",
"gadgets",
"broadband",
"interested",
"websites",
"computing",
"and",
"tech"
],
[
"and",
"from",
"featuring",
"according",
"also",
"oscars",
"awards",
"regarding",
"starring",
"actors"
],
[
"running",
"promising",
"athletes",
"competing",
"olympics",
"athletics",
"qualifying",
"results",
"competitors",
"olympic"
],
[
"sports",
"matches",
"tennis",
"wimbledon",
"rankings",
"results",
"sport",
"competitors",
"competing",
"winning"
]
] | 1,105.016769 | intfloat/e5-large-v2 | 0.885714 | -0.280193 | 0.223161 | 0.909929 |
BBC News | CombinedTM | 42 | 10 | [
[
"half",
"back",
"wales",
"coach",
"side",
"we",
"ball",
"after",
"game",
"try"
],
[
"executive",
"profits",
"shares",
"corporate",
"sector",
"private",
"firm",
"index",
"costs",
"india"
],
[
"in",
"her",
"was",
"for",
"she",
"best",
"who",
"his",
"won",
"at"
],
[
"search",
"people",
"users",
"microsoft",
"software",
"phone",
"mobile",
"data",
"will",
"internet"
],
[
"gold",
"top",
"white",
"indoor",
"number",
"event",
"round",
"award",
"champion",
"norwich"
],
[
"afford",
"mail",
"achieved",
"readers",
"often",
"innocent",
"makers",
"cool",
"person",
"living"
],
[
"oil",
"bank",
"dollar",
"year",
"yukos",
"china",
"rates",
"us",
"economy",
"growth"
],
[
"the",
"it",
"of",
"be",
"is",
"to",
"in",
"for",
"and",
"that"
],
[
"would",
"party",
"labour",
"he",
"said",
"election",
"mr",
"government",
"chancellor",
"prime"
],
[
"written",
"bad",
"introduce",
"contribution",
"opinion",
"innocent",
"learning",
"31",
"person",
"remember"
]
] | 302.801134 | intfloat/e5-large-v2 | 0.96 | -0.097114 | 0.171554 | 0.815293 |
BBC News | CombinedTM | 42 | 20 | [
[
"security",
"microsoft",
"software",
"users",
"windows",
"virus",
"mail",
"calls",
"broadband",
"program"
],
[
"mr",
"his",
"was",
"blair",
"party",
"election",
"said",
"prime",
"he",
"labour"
],
[
"prices",
"december",
"us",
"bank",
"interest",
"rates",
"growth",
"rate",
"year",
"rise"
],
[
"workers",
"howard",
"pension",
"unions",
"mr",
"choice",
"election",
"tories",
"age",
"public"
],
[
"to",
"the",
"we",
"in",
"and",
"have",
"by",
"of",
"be",
"that"
],
[
"lord",
"cards",
"said",
"would",
"id",
"bill",
"law",
"plans",
"police",
"government"
],
[
"arsenal",
"league",
"game",
"playing",
"club",
"play",
"injury",
"england",
"team",
"robinson"
],
[
"leave",
"collins",
"birmingham",
"arrested",
"competing",
"george",
"charge",
"person",
"steve",
"shape"
],
[
"necessary",
"stations",
"fair",
"once",
"yet",
"communications",
"vital",
"fresh",
"skills",
"employers"
],
[
"ball",
"first",
"after",
"second",
"lead",
"minutes",
"try",
"side",
"half",
"france"
],
[
"sites",
"said",
"file",
"files",
"hunting",
"site",
"information",
"traffic",
"they",
"spam"
],
[
"gold",
"indoor",
"believes",
"olympic",
"quite",
"white",
"open",
"champion",
"event",
"federation"
],
[
"of",
"to",
"spanish",
"that",
"and",
"is",
"it",
"the",
"hop",
"are"
],
[
"are",
"and",
"networks",
"3g",
"of",
"to",
"technology",
"is",
"data",
"speeds"
],
[
"comedy",
"office",
"chart",
"film",
"band",
"album",
"oscar",
"director",
"awards",
"award"
],
[
"sell",
"company",
"airlines",
"board",
"shares",
"executive",
"airline",
"takeover",
"deal",
"indian"
],
[
"yukos",
"its",
"oil",
"gazprom",
"london",
"russian",
"russia",
"deutsche",
"bankruptcy",
"has"
],
[
"trade",
"growth",
"figures",
"2004",
"exports",
"rose",
"strong",
"january",
"sales",
"profits"
],
[
"mobiles",
"tv",
"music",
"uk",
"will",
"games",
"phones",
"which",
"more",
"sony"
],
[
"was",
"she",
"who",
"her",
"in",
"his",
"lions",
"best",
"for",
"he"
]
] | 413.882725 | intfloat/e5-large-v2 | 0.905 | -0.033729 | 0.173099 | 0.815854 |
BBC News | CombinedTM | 42 | 30 | [
[
"russia",
"auction",
"yukos",
"oil",
"bankruptcy",
"gazprom",
"russian",
"court",
"state",
"unit"
],
[
"band",
"ceremony",
"male",
"winner",
"gold",
"award",
"prize",
"named",
"international",
"francis"
],
[
"france",
"nations",
"six",
"team",
"wilkinson",
"england",
"squad",
"robinson",
"coach",
"referee"
],
[
"brown",
"taxes",
"chancellor",
"labour",
"schools",
"tories",
"election",
"cuts",
"howard",
"tax"
],
[
"social",
"indian",
"banks",
"sell",
"lending",
"potential",
"shares",
"loans",
"longer",
"airlines"
],
[
"reform",
"agreement",
"send",
"1999",
"begin",
"colleagues",
"heard",
"resigned",
"labels",
"together"
],
[
"minister",
"prime",
"brown",
"book",
"mr",
"blair",
"cabinet",
"he",
"had",
"allies"
],
[
"men",
"mark",
"was",
"she",
"olympic",
"her",
"in",
"race",
"title",
"qualifying"
],
[
"dollar",
"in",
"korea",
"south",
"economic",
"economy",
"currency",
"countries",
"construction",
"budget"
],
[
"of",
"liverpool",
"football",
"stadium",
"he",
"club",
"and",
"the",
"steven",
"that"
],
[
"industries",
"coalition",
"payments",
"winter",
"contracts",
"gives",
"nor",
"energy",
"promise",
"poverty"
],
[
"awards",
"actor",
"best",
"nominated",
"film",
"actress",
"role",
"award",
"for",
"aviator"
],
[
"profits",
"social",
"latin",
"shares",
"sell",
"helped",
"exports",
"strong",
"america",
"forecasts"
],
[
"lord",
"id",
"lords",
"said",
"law",
"would",
"murder",
"cards",
"advice",
"human"
],
[
"growth",
"month",
"prices",
"figures",
"economist",
"december",
"2004",
"year",
"housing",
"rates"
],
[
"to",
"network",
"and",
"of",
"definition",
"on",
"will",
"technology",
"3g",
"the"
],
[
"city",
"league",
"arsenal",
"west",
"manchester",
"rugby",
"injury",
"united",
"football",
"real"
],
[
"tv",
"broadcasting",
"programmes",
"said",
"jupiter",
"mobiles",
"broadband",
"music",
"digital",
"gartner"
],
[
"penalty",
"gara",
"after",
"shot",
"minute",
"half",
"break",
"second",
"minutes",
"ball"
],
[
"drugs",
"charged",
"happened",
"doping",
"suspended",
"community",
"assault",
"banned",
"clean",
"appeared"
],
[
"number",
"box",
"starring",
"office",
"film",
"chart",
"films",
"oscar",
"sequel",
"album"
],
[
"programs",
"software",
"security",
"users",
"microsoft",
"viruses",
"windows",
"spam",
"search",
"mail"
],
[
"are",
"they",
"sites",
"to",
"file",
"said",
"attacks",
"that",
"traffic",
"students"
],
[
"his",
"was",
"he",
"silk",
"mr",
"election",
"party",
"said",
"labour",
"the"
],
[
"prescott",
"correspondent",
"pension",
"hunting",
"workers",
"sector",
"police",
"000",
"unions",
"union"
],
[
"the",
"and",
"is",
"it",
"hip",
"spanish",
"of",
"hop",
"radio",
"rap"
],
[
"stock",
"exchange",
"china",
"deutsche",
"london",
"takeover",
"shareholders",
"its",
"offer",
"bid"
],
[
"seed",
"open",
"champion",
"cross",
"championships",
"indoor",
"round",
"australian",
"federer",
"edinburgh"
],
[
"handheld",
"text",
"sony",
"gaming",
"psp",
"playstation",
"console",
"nintendo",
"xbox",
"titles"
],
[
"that",
"threat",
"to",
"of",
"in",
"the",
"and",
"minimum",
"agree",
"be"
]
] | 330.504946 | intfloat/e5-large-v2 | 0.893333 | -0.006361 | 0.157415 | 0.858588 |
BBC News | CombinedTM | 42 | 40 | [
[
"to",
"that",
"the",
"of",
"it",
"list",
"and",
"is",
"be",
"modern"
],
[
"jones",
"france",
"thomas",
"after",
"minutes",
"half",
"wales",
"kick",
"from",
"two"
],
[
"budget",
"foreign",
"economic",
"spending",
"imf",
"will",
"reform",
"aids",
"bush",
"be"
],
[
"audiences",
"30",
"britons",
"supreme",
"highest",
"flight",
"single",
"festival",
"died",
"guardian"
],
[
"secret",
"resigned",
"arrested",
"fight",
"lives",
"commit",
"green",
"documentary",
"story",
"wonderful"
],
[
"in",
"coach",
"woodward",
"nations",
"lions",
"their",
"cup",
"eddie",
"scotland",
"rugby"
],
[
"attacks",
"sites",
"to",
"centres",
"traffic",
"are",
"attack",
"call",
"site",
"criminals"
],
[
"yen",
"airlines",
"india",
"indian",
"japan",
"japanese",
"banks",
"15",
"poverty",
"development"
],
[
"his",
"jewish",
"blair",
"mr",
"poster",
"he",
"apologise",
"minister",
"ambition",
"prime"
],
[
"file",
"said",
"sharing",
"files",
"peer",
"networks",
"sites",
"were",
"version",
"that"
],
[
"costs",
"profit",
"its",
"euros",
"increased",
"profits",
"largest",
"china",
"deal",
"heineken"
],
[
"trade",
"us",
"dollar",
"fed",
"china",
"deficit",
"exports",
"reserve",
"economic",
"currency"
],
[
"murder",
"law",
"judges",
"human",
"lord",
"rights",
"lords",
"bill",
"said",
"justice"
],
[
"oil",
"its",
"gazprom",
"india",
"yukos",
"russia",
"company",
"russian",
"standard",
"production"
],
[
"tories",
"would",
"kennedy",
"labour",
"students",
"plans",
"howard",
"election",
"police",
"commissioner"
],
[
"company",
"products",
"earnings",
"profits",
"poor",
"shares",
"needs",
"drug",
"insurance",
"stock"
],
[
"fixed",
"calls",
"broadband",
"uk",
"bt",
"service",
"speed",
"via",
"million",
"internet"
],
[
"dvds",
"definition",
"games",
"format",
"generation",
"xbox",
"dvd",
"console",
"sony",
"ray"
],
[
"draw",
"west",
"callaghan",
"cap",
"everton",
"obvious",
"white",
"edinburgh",
"road",
"saturday"
],
[
"album",
"box",
"chart",
"office",
"number",
"success",
"weekend",
"single",
"place",
"band"
],
[
"rise",
"january",
"slowdown",
"interest",
"figures",
"rates",
"bank",
"month",
"housing",
"december"
],
[
"team",
"game",
"out",
"we",
"very",
"half",
"referee",
"robinson",
"england",
"nations"
],
[
"queen",
"elvis",
"episode",
"show",
"alleged",
"appeared",
"story",
"star",
"singer",
"lawyer"
],
[
"they",
"to",
"marketing",
"voting",
"rules",
"if",
"intentions",
"premium",
"stop",
"telephone"
],
[
"workers",
"age",
"pension",
"public",
"local",
"correspondent",
"prescott",
"000",
"union",
"sector"
],
[
"the",
"minimum",
"of",
"to",
"lord",
"mr",
"said",
"in",
"he",
"trial"
],
[
"stone",
"her",
"black",
"in",
"the",
"she",
"music",
"soul",
"of",
"term"
],
[
"offer",
"bid",
"deutsche",
"exchange",
"london",
"stock",
"shareholders",
"takeover",
"talks",
"compensation"
],
[
"greek",
"contract",
"tests",
"madrid",
"drugs",
"club",
"arsenal",
"athens",
"doping",
"missing"
],
[
"commons",
"super",
"charged",
"together",
"gives",
"ban",
"supporters",
"merger",
"unions",
"behaviour"
],
[
"after",
"henman",
"set",
"roddick",
"first",
"was",
"celtic",
"final",
"second",
"ball"
],
[
"his",
"silk",
"election",
"mr",
"he",
"party",
"book",
"brown",
"stand",
"robert"
],
[
"actor",
"awards",
"film",
"nominated",
"ceremony",
"nominees",
"best",
"oscar",
"actress",
"award"
],
[
"phones",
"mobiles",
"cameras",
"services",
"mobile",
"broadcasting",
"battery",
"gartner",
"handsets",
"camera"
],
[
"spanish",
"hip",
"rap",
"is",
"hop",
"radio",
"mini",
"mac",
"it",
"and"
],
[
"tories",
"cuts",
"howard",
"tax",
"taxes",
"election",
"chancellor",
"labour",
"cut",
"tory"
],
[
"3g",
"networks",
"speeds",
"second",
"to",
"operators",
"faster",
"of",
"and",
"mobile"
],
[
"users",
"virus",
"programs",
"spam",
"linux",
"mail",
"search",
"software",
"windows",
"microsoft"
],
[
"form",
"round",
"champion",
"indoor",
"australian",
"gold",
"holmes",
"open",
"world",
"seed"
],
[
"his",
"he",
"club",
"chelsea",
"smith",
"was",
"players",
"sport",
"tennis",
"liverpool"
]
] | 427.362812 | intfloat/e5-large-v2 | 0.875 | -0.028478 | 0.154016 | 0.854466 |
BBC News | CombinedTM | 42 | 50 | [
[
"in",
"title",
"seconds",
"mark",
"her",
"60m",
"european",
"indoor",
"jump",
"qualifying"
],
[
"visual",
"mini",
"and",
"it",
"mac",
"computer",
"revolution",
"is",
"the",
"project"
],
[
"rates",
"us",
"interest",
"economists",
"manufacturing",
"bank",
"dollar",
"slowdown",
"economy",
"fed"
],
[
"traffic",
"websites",
"spam",
"attack",
"sites",
"data",
"attacks",
"site",
"spammers",
"security"
],
[
"broadband",
"uk",
"million",
"bt",
"calls",
"text",
"people",
"fast",
"net",
"service"
],
[
"she",
"should",
"failed",
"drugs",
"situation",
"police",
"parties",
"there",
"make",
"if"
],
[
"blame",
"striker",
"supporters",
"manager",
"manchester",
"everton",
"fa",
"west",
"trouble",
"behaviour"
],
[
"will",
"parents",
"account",
"children",
"child",
"be",
"federal",
"tax",
"budget",
"fund"
],
[
"suggest",
"1980s",
"elections",
"age",
"carried",
"electoral",
"psp",
"below",
"adults",
"suggests"
],
[
"khodorkovsky",
"russia",
"court",
"its",
"yukos",
"oil",
"russian",
"bankruptcy",
"gazprom",
"unit"
],
[
"of",
"steve",
"to",
"the",
"steven",
"stadium",
"new",
"liverpool",
"we",
"door"
],
[
"high",
"generation",
"dvds",
"dvd",
"ray",
"technology",
"toshiba",
"format",
"games",
"data"
],
[
"lords",
"ruling",
"human",
"murder",
"rights",
"court",
"law",
"said",
"laws",
"trial"
],
[
"silk",
"he",
"eu",
"party",
"robert",
"leadership",
"mr",
"independence",
"vote",
"was"
],
[
"howard",
"taxes",
"tory",
"cuts",
"tax",
"would",
"labour",
"tories",
"stability",
"plans"
],
[
"am",
"arsenal",
"madrid",
"real",
"club",
"contract",
"league",
"added",
"playing",
"me"
],
[
"sold",
"sony",
"death",
"show",
"psp",
"stuff",
"day",
"festival",
"shows",
"christmas"
],
[
"potential",
"financial",
"shares",
"company",
"banks",
"sec",
"stock",
"agreed",
"firms",
"investors"
],
[
"cinema",
"weekend",
"film",
"office",
"box",
"starring",
"1m",
"comedy",
"films",
"5m"
],
[
"services",
"handsets",
"gartner",
"camera",
"broadcasting",
"cameras",
"phones",
"will",
"mobiles",
"digital"
],
[
"xbox",
"halo",
"wales",
"game",
"games",
"nintendo",
"grand",
"console",
"san",
"titles"
],
[
"to",
"google",
"net",
"internet",
"regulation",
"content",
"thinks",
"is",
"it",
"search"
],
[
"australian",
"set",
"roddick",
"davis",
"really",
"federer",
"roger",
"match",
"serve",
"felt"
],
[
"2004",
"figures",
"december",
"sales",
"unemployment",
"growth",
"retail",
"previous",
"rose",
"rate"
],
[
"it",
"that",
"to",
"and",
"robbie",
"is",
"the",
"of",
"25",
"but"
],
[
"players",
"sport",
"me",
"club",
"lions",
"team",
"he",
"admitted",
"england",
"chelsea"
],
[
"aviator",
"director",
"baby",
"performance",
"award",
"actress",
"named",
"film",
"won",
"nominees"
],
[
"arrested",
"mps",
"members",
"leicester",
"charged",
"failing",
"together",
"west",
"commit",
"fa"
],
[
"to",
"premium",
"argument",
"numbers",
"they",
"are",
"companies",
"stop",
"calls",
"and"
],
[
"bosses",
"2007",
"vital",
"delegates",
"engines",
"6bn",
"loss",
"pushing",
"tape",
"longer"
],
[
"george",
"foot",
"cool",
"wars",
"wrong",
"character",
"highly",
"novel",
"informed",
"fantastic"
],
[
"programs",
"software",
"microsoft",
"virus",
"linux",
"windows",
"users",
"mail",
"source",
"ibm"
],
[
"sell",
"takeover",
"friday",
"fund",
"stations",
"shareholders",
"industries",
"offer",
"buy",
"agreed"
],
[
"saw",
"strong",
"profit",
"profits",
"heineken",
"western",
"china",
"sales",
"rise",
"europe"
],
[
"brown",
"labour",
"mr",
"role",
"chancellor",
"milburn",
"minister",
"sullivan",
"prime",
"blair"
],
[
"bid",
"deal",
"deutsche",
"shareholders",
"group",
"london",
"exchange",
"card",
"takeover",
"offer"
],
[
"number",
"gold",
"chart",
"album",
"indoor",
"elvis",
"olympic",
"record",
"60m",
"medal"
],
[
"he",
"mr",
"jewish",
"blunkett",
"his",
"blair",
"howard",
"poster",
"had",
"image"
],
[
"magazine",
"the",
"in",
"100",
"and",
"of",
"mobile",
"gadget",
"swiss",
"gadgets"
],
[
"said",
"files",
"mac",
"system",
"systems",
"basic",
"peer",
"sharing",
"file",
"apple"
],
[
"local",
"correspondent",
"sector",
"pension",
"prescott",
"union",
"choice",
"unions",
"workers",
"government"
],
[
"tax",
"labour",
"voters",
"on",
"his",
"election",
"he",
"spending",
"taxes",
"outcome"
],
[
"theatre",
"leigh",
"nominated",
"nomination",
"role",
"she",
"film",
"category",
"academy",
"band"
],
[
"half",
"robinson",
"six",
"side",
"hodgson",
"england",
"france",
"wales",
"we",
"cardiff"
],
[
"written",
"lord",
"military",
"falconer",
"advice",
"judge",
"said",
"was",
"attorney",
"inquiry"
],
[
"foreign",
"councils",
"homes",
"housing",
"plans",
"property",
"would",
"powers",
"budget",
"straw"
],
[
"edinburgh",
"glasgow",
"simon",
"round",
"taylor",
"italy",
"leeds",
"borders",
"beat",
"murray"
],
[
"penalty",
"minutes",
"ball",
"gara",
"after",
"minute",
"goal",
"lead",
"break",
"ireland"
],
[
"helped",
"fiscal",
"products",
"area",
"asia",
"manufacturing",
"indian",
"gas",
"factory",
"exports"
],
[
"charles",
"uk",
"was",
"film",
"bafta",
"the",
"of",
"and",
"in",
"vera"
]
] | 388.248686 | intfloat/e5-large-v2 | 0.86 | -0.025484 | 0.157735 | 0.864229 |
BBC News | ZeroShotTM | 42 | 10 | [
[
"of",
"and",
"to",
"as",
"that",
"are",
"is",
"it",
"the",
"tv"
],
[
"2004",
"growth",
"shares",
"profits",
"deal",
"costs",
"stock",
"exports",
"strong",
"higher"
],
[
"awards",
"stars",
"bond",
"comedy",
"actor",
"category",
"director",
"movie",
"aviator",
"leigh"
],
[
"said",
"mr",
"howard",
"government",
"brown",
"he",
"would",
"chancellor",
"prime",
"election"
],
[
"of",
"the",
"in",
"and",
"to",
"that",
"is",
"it",
"we",
"for"
],
[
"side",
"second",
"after",
"nations",
"ireland",
"match",
"half",
"game",
"england",
"six"
],
[
"economic",
"us",
"economy",
"growth",
"december",
"prices",
"yukos",
"oil",
"dollar",
"interest"
],
[
"said",
"they",
"sites",
"be",
"to",
"attacks",
"spam",
"would",
"site",
"rights"
],
[
"afford",
"anybody",
"keep",
"eye",
"comes",
"towards",
"cards",
"seems",
"medal",
"everybody"
],
[
"software",
"programs",
"free",
"allow",
"cards",
"aimed",
"protect",
"peer",
"piracy",
"launched"
]
] | 287.512351 | intfloat/e5-large-v2 | 0.88 | -0.048381 | 0.184823 | 0.791482 |
BBC News | ZeroShotTM | 42 | 20 | [
[
"of",
"that",
"to",
"for",
"be",
"firms",
"said",
"sites",
"by",
"are"
],
[
"two",
"minutes",
"ball",
"game",
"half",
"england",
"side",
"try",
"after",
"second"
],
[
"figures",
"rise",
"sales",
"economist",
"rates",
"prices",
"growth",
"november",
"economy",
"housing"
],
[
"russian",
"court",
"yukos",
"state",
"bankruptcy",
"russia",
"gazprom",
"oil",
"its",
"us"
],
[
"deutsche",
"company",
"stock",
"shares",
"talks",
"firm",
"executive",
"bid",
"deal",
"shareholders"
],
[
"madrid",
"club",
"arsenal",
"league",
"scotland",
"season",
"cup",
"football",
"return",
"edinburgh"
],
[
"said",
"he",
"plans",
"police",
"would",
"law",
"mr",
"was",
"local",
"be"
],
[
"and",
"definition",
"content",
"video",
"dvd",
"mac",
"tv",
"is",
"mini",
"it"
],
[
"he",
"would",
"mr",
"said",
"had",
"his",
"police",
"was",
"law",
"public"
],
[
"film",
"box",
"comedy",
"office",
"number",
"award",
"baby",
"album",
"actress",
"awards"
],
[
"eye",
"described",
"size",
"consider",
"fines",
"every",
"jackson",
"protests",
"individual",
"giving"
],
[
"service",
"million",
"broadband",
"microsoft",
"software",
"users",
"free",
"calls",
"uk",
"bt"
],
[
"brown",
"labour",
"mr",
"blair",
"election",
"howard",
"he",
"leader",
"chancellor",
"prime"
],
[
"died",
"whose",
"actor",
"prince",
"live",
"nominated",
"awards",
"oscar",
"singer",
"day"
],
[
"tennis",
"championships",
"return",
"open",
"holmes",
"ll",
"seed",
"australian",
"000m",
"title"
],
[
"of",
"the",
"to",
"and",
"that",
"we",
"is",
"it",
"in",
"have"
],
[
"song",
"in",
"best",
"the",
"and",
"it",
"for",
"list",
"25",
"of"
],
[
"wales",
"was",
"coach",
"ireland",
"victory",
"six",
"england",
"his",
"after",
"jones"
],
[
"shares",
"profits",
"stock",
"company",
"market",
"china",
"expected",
"investment",
"growth",
"term"
],
[
"schools",
"servants",
"age",
"elections",
"voting",
"shown",
"education",
"episode",
"effectively",
"electoral"
]
] | 307.65923 | intfloat/e5-large-v2 | 0.835 | -0.008999 | 0.171164 | 0.817988 |
BBC News | ZeroShotTM | 42 | 30 | [
[
"children",
"around",
"box",
"moving",
"mail",
"itunes",
"receive",
"release",
"across",
"aid"
],
[
"victory",
"in",
"nations",
"olympic",
"six",
"season",
"was",
"world",
"indoor",
"championship"
],
[
"first",
"after",
"two",
"france",
"try",
"half",
"minutes",
"ireland",
"ball",
"lead"
],
[
"award",
"comedy",
"film",
"won",
"best",
"director",
"actress",
"awards",
"aviator",
"nominated"
],
[
"company",
"firm",
"cost",
"euros",
"profit",
"source",
"profits",
"shares",
"2bn",
"executive"
],
[
"to",
"in",
"the",
"and",
"of",
"for",
"that",
"is",
"we",
"but"
],
[
"said",
"plans",
"labour",
"he",
"was",
"mr",
"would",
"had",
"party",
"be"
],
[
"brown",
"said",
"mr",
"tories",
"would",
"plans",
"shadow",
"government",
"bill",
"howard"
],
[
"unions",
"gas",
"private",
"needs",
"pensions",
"agreed",
"merger",
"al",
"details",
"agreement"
],
[
"buy",
"euros",
"shareholders",
"stock",
"company",
"exchange",
"boost",
"deutsche",
"firm",
"commission"
],
[
"place",
"gold",
"olympic",
"medal",
"indoor",
"great",
"top",
"race",
"topped",
"third"
],
[
"dvd",
"mini",
"and",
"that",
"it",
"is",
"you",
"to",
"of",
"mac"
],
[
"law",
"would",
"said",
"government",
"yukos",
"court",
"budget",
"mr",
"should",
"bill"
],
[
"attacks",
"sites",
"said",
"traffic",
"spam",
"websites",
"they",
"to",
"peer",
"information"
],
[
"call",
"doping",
"ban",
"entirely",
"women",
"age",
"white",
"drugs",
"seriously",
"everyone"
],
[
"found",
"windows",
"users",
"used",
"internet",
"software",
"microsoft",
"more",
"computer",
"people"
],
[
"prices",
"december",
"sales",
"growth",
"rise",
"month",
"consumer",
"retail",
"us",
"economy"
],
[
"he",
"mr",
"blair",
"labour",
"howard",
"said",
"had",
"party",
"war",
"was"
],
[
"beginning",
"spain",
"course",
"defend",
"fire",
"size",
"students",
"learning",
"cap",
"suspended"
],
[
"japan",
"psp",
"nintendo",
"america",
"text",
"million",
"linux",
"five",
"sony",
"handheld"
],
[
"will",
"cell",
"mobile",
"tv",
"mobiles",
"be",
"services",
"gartner",
"phones",
"3g"
],
[
"robbie",
"song",
"25",
"of",
"in",
"it",
"list",
"the",
"and",
"is"
],
[
"the",
"in",
"be",
"budget",
"spending",
"economic",
"to",
"by",
"of",
"tax"
],
[
"football",
"training",
"glasgow",
"madrid",
"squad",
"edinburgh",
"fit",
"murray",
"stadium",
"rugby"
],
[
"his",
"him",
"you",
"he",
"me",
"team",
"players",
"ve",
"scottish",
"club"
],
[
"and",
"of",
"the",
"they",
"have",
"to",
"be",
"it",
"are",
"that"
],
[
"howard",
"leader",
"mr",
"mp",
"jewish",
"blair",
"he",
"party",
"stand",
"labour"
],
[
"year",
"broadband",
"uk",
"2005",
"prices",
"than",
"total",
"growth",
"market",
"million"
],
[
"deutsche",
"oil",
"yukos",
"its",
"gazprom",
"russia",
"russian",
"bankruptcy",
"court",
"london"
],
[
"coach",
"side",
"league",
"game",
"match",
"club",
"cup",
"half",
"football",
"win"
]
] | 332.907457 | intfloat/e5-large-v2 | 0.77 | -0.019583 | 0.187925 | 0.811652 |
BBC News | ZeroShotTM | 42 | 40 | [
[
"musical",
"director",
"comedy",
"sideways",
"starring",
"box",
"award",
"office",
"named",
"critics"
],
[
"uk",
"spam",
"found",
"more",
"people",
"or",
"broadband",
"bt",
"service",
"users"
],
[
"2004",
"profits",
"euros",
"figures",
"strong",
"rose",
"overall",
"exports",
"growth",
"january"
],
[
"net",
"sites",
"search",
"security",
"to",
"regulation",
"systems",
"traffic",
"content",
"they"
],
[
"court",
"charges",
"tried",
"guilty",
"ms",
"athletes",
"legal",
"lawyer",
"case",
"claim"
],
[
"yukos",
"oil",
"gazprom",
"its",
"russian",
"unit",
"us",
"bankruptcy",
"china",
"deutsche"
],
[
"he",
"tony",
"said",
"mr",
"was",
"advice",
"election",
"blair",
"prime",
"question"
],
[
"film",
"for",
"best",
"awards",
"night",
"actress",
"nominated",
"who",
"won",
"supporting"
],
[
"mourinho",
"white",
"debate",
"fa",
"answer",
"ruled",
"west",
"happened",
"draw",
"cancelled"
],
[
"the",
"in",
"to",
"liverpool",
"and",
"is",
"that",
"of",
"it",
"have"
],
[
"interest",
"prices",
"economy",
"sales",
"rates",
"rate",
"economist",
"december",
"growth",
"fall"
],
[
"of",
"the",
"song",
"it",
"25",
"robbie",
"to",
"that",
"and",
"songs"
],
[
"deal",
"group",
"agreed",
"president",
"workers",
"executive",
"state",
"sec",
"negotiations",
"foreign"
],
[
"nations",
"in",
"holmes",
"zealand",
"six",
"wales",
"title",
"season",
"cup",
"championships"
],
[
"club",
"players",
"team",
"sport",
"boss",
"arsenal",
"squad",
"told",
"england",
"madrid"
],
[
"of",
"they",
"he",
"is",
"his",
"chelsea",
"it",
"club",
"with",
"scottish"
],
[
"deal",
"buy",
"social",
"stock",
"airlines",
"shares",
"share",
"sell",
"stake",
"raise"
],
[
"band",
"favourite",
"stars",
"television",
"show",
"director",
"hill",
"star",
"series",
"live"
],
[
"murray",
"leeds",
"edinburgh",
"glasgow",
"ll",
"students",
"rugby",
"borders",
"playing",
"scotland"
],
[
"strong",
"sales",
"profits",
"fall",
"china",
"growth",
"rise",
"jobs",
"market",
"2004"
],
[
"debut",
"record",
"compete",
"again",
"gold",
"greatest",
"olympic",
"stars",
"world",
"her"
],
[
"comedy",
"aviator",
"baby",
"prize",
"favourite",
"oscar",
"7m",
"among",
"office",
"sideways"
],
[
"we",
"robinson",
"hodgson",
"england",
"six",
"half",
"goal",
"lewsey",
"side",
"ireland"
],
[
"plans",
"bill",
"tax",
"lord",
"law",
"would",
"laws",
"review",
"said",
"local"
],
[
"the",
"of",
"silk",
"to",
"and",
"in",
"party",
"voting",
"he",
"is"
],
[
"security",
"virus",
"programs",
"windows",
"anti",
"microsoft",
"mails",
"software",
"mail",
"itself"
],
[
"mr",
"blair",
"labour",
"howard",
"he",
"prime",
"mp",
"party",
"tory",
"leader"
],
[
"said",
"mr",
"they",
"tory",
"would",
"he",
"blair",
"election",
"labour",
"howard"
],
[
"indoor",
"final",
"champion",
"race",
"win",
"olympic",
"seed",
"000m",
"jump",
"federer"
],
[
"children",
"countries",
"debt",
"finance",
"child",
"of",
"economic",
"be",
"health",
"are"
],
[
"it",
"definition",
"and",
"mac",
"dvds",
"the",
"dvd",
"mini",
"is",
"revolution"
],
[
"voice",
"users",
"available",
"version",
"linux",
"text",
"mail",
"pages",
"ibm",
"access"
],
[
"gamers",
"nintendo",
"console",
"games",
"are",
"gaming",
"sony",
"will",
"music",
"screens"
],
[
"victory",
"henman",
"second",
"set",
"final",
"roddick",
"title",
"was",
"race",
"his"
],
[
"oil",
"gazprom",
"its",
"korea",
"yukos",
"south",
"deutsche",
"budget",
"russian",
"bush"
],
[
"networks",
"said",
"speeds",
"to",
"3g",
"will",
"technology",
"and",
"operator",
"its"
],
[
"elections",
"workers",
"pension",
"pensions",
"voting",
"employees",
"retirement",
"social",
"commons",
"once"
],
[
"minutes",
"second",
"two",
"side",
"after",
"half",
"ball",
"goal",
"final",
"gara"
],
[
"injury",
"match",
"seed",
"draw",
"champion",
"championships",
"ferguson",
"open",
"wins",
"spaniard"
],
[
"uefa",
"steve",
"together",
"phillips",
"downing",
"chelsea",
"friendly",
"leicester",
"west",
"appears"
]
] | 432.617197 | intfloat/e5-large-v2 | 0.765 | -0.03025 | 0.162515 | 0.843954 |
BBC News | ZeroShotTM | 42 | 50 | [
[
"party",
"howard",
"mp",
"mr",
"leader",
"blair",
"apologise",
"jewish",
"labour",
"trust"
],
[
"issues",
"mp",
"seats",
"mourinho",
"chelsea",
"denies",
"denied",
"conservative",
"wednesday",
"added"
],
[
"final",
"my",
"championships",
"top",
"holmes",
"her",
"60m",
"indoor",
"olympic",
"race"
],
[
"party",
"silk",
"robert",
"to",
"parties",
"leadership",
"the",
"of",
"and",
"he"
],
[
"thinks",
"and",
"to",
"trusted",
"you",
"hardware",
"it",
"your",
"3g",
"that"
],
[
"sequel",
"halo",
"game",
"100",
"gadgets",
"gadget",
"magazine",
"xbox",
"list",
"consoles"
],
[
"adults",
"promoting",
"male",
"rating",
"applied",
"area",
"ways",
"moves",
"promote",
"suffer"
],
[
"uncertain",
"students",
"afford",
"approach",
"positive",
"collins",
"chambers",
"glasgow",
"united",
"welsh"
],
[
"learning",
"peer",
"you",
"games",
"it",
"help",
"is",
"game",
"they",
"search"
],
[
"match",
"final",
"seed",
"beat",
"win",
"open",
"round",
"federer",
"set",
"victory"
],
[
"gartner",
"will",
"mobiles",
"screens",
"cameras",
"definition",
"camera",
"generation",
"tv",
"broadcasting"
],
[
"it",
"spanish",
"the",
"hop",
"and",
"rap",
"hip",
"of",
"on",
"to"
],
[
"england",
"goal",
"barkley",
"half",
"graham",
"ireland",
"six",
"lead",
"ball",
"france"
],
[
"euros",
"shares",
"profits",
"safety",
"recall",
"indian",
"vehicles",
"trade",
"grew",
"gas"
],
[
"prices",
"us",
"rates",
"rate",
"economist",
"bank",
"interest",
"dollar",
"inflation",
"rise"
],
[
"book",
"he",
"mr",
"his",
"allies",
"said",
"cabinet",
"blair",
"minister",
"had"
],
[
"blame",
"white",
"targets",
"murray",
"collins",
"charge",
"heard",
"cancelled",
"students",
"happened"
],
[
"bush",
"korea",
"gazprom",
"india",
"oil",
"budget",
"south",
"economic",
"economy",
"its"
],
[
"lords",
"rights",
"yukos",
"foreign",
"clarke",
"would",
"human",
"said",
"court",
"government"
],
[
"died",
"singer",
"star",
"death",
"show",
"attended",
"woman",
"queen",
"killed",
"acts"
],
[
"title",
"was",
"roddick",
"olympic",
"final",
"in",
"men",
"after",
"fourth",
"her"
],
[
"awards",
"director",
"prize",
"actor",
"aviator",
"award",
"film",
"oscar",
"best",
"scorsese"
],
[
"stock",
"shares",
"exchange",
"euros",
"china",
"profits",
"unit",
"investors",
"costs",
"investment"
],
[
"club",
"stadium",
"of",
"liverpool",
"is",
"the",
"that",
"to",
"he",
"and"
],
[
"oil",
"government",
"workers",
"prescott",
"us",
"talks",
"its",
"energy",
"economic",
"economy"
],
[
"film",
"films",
"stars",
"warner",
"hits",
"director",
"hollywood",
"book",
"release",
"dvd"
],
[
"lions",
"woodward",
"zealand",
"nations",
"england",
"six",
"rugby",
"wales",
"coach",
"referee"
],
[
"shareholder",
"deutsche",
"fund",
"shares",
"company",
"shareholders",
"board",
"exchange",
"projects",
"takeover"
],
[
"programs",
"microsoft",
"software",
"users",
"virus",
"mail",
"windows",
"google",
"security",
"free"
],
[
"social",
"pensions",
"commons",
"message",
"science",
"unions",
"gives",
"super",
"energy",
"pension"
],
[
"united",
"manchester",
"triple",
"sheffield",
"ferguson",
"ll",
"everton",
"draw",
"west",
"open"
],
[
"debt",
"children",
"imf",
"fund",
"of",
"are",
"to",
"the",
"be",
"and"
],
[
"group",
"airline",
"shareholders",
"deutsche",
"exchange",
"financial",
"london",
"shareholder",
"company",
"talks"
],
[
"rules",
"marketing",
"law",
"police",
"watchdog",
"commissioner",
"information",
"id",
"hunting",
"hunt"
],
[
"younger",
"workers",
"unions",
"family",
"matter",
"extra",
"moves",
"ambitions",
"prince",
"35"
],
[
"be",
"police",
"would",
"plans",
"said",
"asylum",
"home",
"id",
"local",
"scheme"
],
[
"mr",
"said",
"he",
"labour",
"election",
"had",
"party",
"his",
"was",
"britain"
],
[
"we",
"half",
"try",
"minutes",
"side",
"back",
"goals",
"after",
"game",
"ball"
],
[
"sharing",
"firms",
"attacks",
"systems",
"sites",
"spam",
"websites",
"content",
"file",
"files"
],
[
"programs",
"mails",
"blue",
"security",
"software",
"spam",
"viruses",
"mail",
"via",
"available"
],
[
"wilkinson",
"we",
"england",
"arsenal",
"team",
"sport",
"club",
"chelsea",
"robinson",
"playing"
],
[
"brown",
"tax",
"taxes",
"mr",
"chancellor",
"government",
"lord",
"cuts",
"judge",
"civil"
],
[
"best",
"25",
"and",
"voted",
"of",
"the",
"music",
"songs",
"song",
"robbie"
],
[
"actress",
"film",
"best",
"night",
"award",
"aviator",
"awards",
"nominated",
"won",
"charles"
],
[
"place",
"album",
"earned",
"tour",
"1m",
"chart",
"5m",
"debut",
"9m",
"4m"
],
[
"would",
"military",
"conference",
"howard",
"mr",
"advice",
"answer",
"said",
"he",
"attorney"
],
[
"december",
"figures",
"growth",
"2004",
"quarter",
"sales",
"housing",
"january",
"economy",
"overall"
],
[
"box",
"yen",
"sony",
"films",
"japan",
"psp",
"office",
"dvd",
"format",
"disney"
],
[
"mobile",
"line",
"bt",
"call",
"uk",
"speeds",
"speed",
"broadband",
"calls",
"fast"
],
[
"china",
"gazprom",
"yukos",
"bankruptcy",
"oil",
"us",
"russian",
"its",
"khodorkovsky",
"auction"
]
] | 384.126599 | intfloat/e5-large-v2 | 0.774 | -0.044072 | 0.158671 | 0.859618 |
ArXiv ML Papers | GMM | 42 | 10 | [
[
"deep",
"neural",
"layer",
"network",
"architectures",
"weights",
"gradient",
"networks",
"optimization",
"parameters"
],
[
"word",
"language",
"attention",
"transformer",
"text",
"languages",
"speech",
"audio",
"speaker",
"task"
],
[
"against",
"attacks",
"privacy",
"adversarial",
"perturbations",
"robustness",
"defense",
"attack",
"robust",
"examples"
],
[
"ml",
"human",
"machine",
"research",
"available",
"been",
"their",
"code",
"techniques",
"it"
],
[
"graph",
"graphs",
"nodes",
"node",
"embedding",
"recommendation",
"user",
"representation",
"fairness",
"prediction"
],
[
"traffic",
"monitoring",
"time",
"forecasting",
"series",
"forecast",
"prediction",
"demand",
"forecasts",
"temporal"
],
[
"segmentation",
"medical",
"patients",
"net",
"images",
"disease",
"clinical",
"cancer",
"health",
"image"
],
[
"rl",
"games",
"policy",
"agent",
"agents",
"reward",
"regret",
"reinforcement",
"policies",
"control"
],
[
"image",
"images",
"generative",
"gan",
"object",
"video",
"visual",
"scene",
"domain",
"dataset"
],
[
"regression",
"convex",
"matrix",
"optimization",
"algorithm",
"linear",
"rank",
"bounds",
"kernel",
"sample"
]
] | 5.233222 | intfloat/e5-large-v2 | 0.96 | 0.017663 | 0.15115 | 0.840513 |
ArXiv ML Papers | GMM | 42 | 20 | [
[
"neural",
"network",
"relu",
"layer",
"networks",
"weights",
"pruning",
"layers",
"deep",
"activation"
],
[
"classifier",
"machine",
"decision",
"fair",
"label",
"ml",
"classification",
"fairness",
"research",
"prediction"
],
[
"defense",
"examples",
"attack",
"against",
"attacks",
"robust",
"robustness",
"adversarial",
"perturbations",
"box"
],
[
"tensor",
"quantum",
"human",
"filter",
"machine",
"al",
"et",
"library",
"implementation",
"behavior"
],
[
"federated",
"communication",
"distributed",
"privacy",
"private",
"server",
"devices",
"local",
"decentralized",
"updates"
],
[
"weather",
"forecast",
"forest",
"ensemble",
"prediction",
"forecasting",
"forecasts",
"boosting",
"day",
"time"
],
[
"medical",
"images",
"cell",
"imaging",
"image",
"segmentation",
"brain",
"shape",
"net",
"labeled"
],
[
"sqrt",
"arm",
"regret",
"algorithm",
"optimal",
"bandit",
"bound",
"bandits",
"reward",
"online"
],
[
"images",
"visual",
"object",
"video",
"objects",
"cnn",
"detection",
"scene",
"3d",
"image"
],
[
"bayesian",
"distribution",
"distributions",
"sample",
"estimator",
"inference",
"gaussian",
"bounds",
"posterior",
"variables"
],
[
"physics",
"differential",
"equation",
"equations",
"systems",
"dynamics",
"parameters",
"neural",
"quantum",
"physical"
],
[
"series",
"deep",
"temporal",
"recommendation",
"speech",
"time",
"audio",
"signals",
"feature",
"eeg"
],
[
"disease",
"medical",
"predict",
"health",
"clinical",
"patients",
"covid",
"19",
"was",
"study"
],
[
"policy",
"rl",
"reinforcement",
"policies",
"reward",
"agent",
"agents",
"environment",
"control",
"action"
],
[
"embeddings",
"node",
"graphs",
"graph",
"nodes",
"embedding",
"representation",
"gnns",
"link",
"networks"
],
[
"attention",
"transformer",
"text",
"language",
"languages",
"words",
"word",
"speech",
"representations",
"natural"
],
[
"optimization",
"regression",
"low",
"linear",
"clustering",
"tensor",
"matrix",
"algorithm",
"rank",
"convex"
],
[
"architecture",
"deep",
"convolutional",
"accuracy",
"memory",
"hardware",
"neural",
"nas",
"image",
"architectures"
],
[
"sensor",
"detect",
"vehicle",
"attacks",
"traffic",
"iot",
"vehicles",
"detection",
"driving",
"internet"
],
[
"generative",
"gans",
"gan",
"image",
"images",
"variational",
"latent",
"generation",
"generator",
"vae"
]
] | 8.562865 | intfloat/e5-large-v2 | 0.895 | 0.031863 | 0.152897 | 0.861517 |
ArXiv ML Papers | GMM | 42 | 30 | [
[
"approximate",
"distributions",
"prior",
"bayesian",
"distribution",
"weights",
"inference",
"variational",
"posterior",
"family"
],
[
"reasoning",
"concepts",
"interpretability",
"explanations",
"explanation",
"black",
"explainable",
"interpretable",
"counterfactual",
"logic"
],
[
"attacks",
"detect",
"driving",
"detection",
"vehicle",
"malware",
"attack",
"iot",
"security",
"video"
],
[
"human",
"code",
"language",
"al",
"machine",
"et",
"library",
"tensor",
"recurrent",
"project"
],
[
"low",
"matrix",
"sparse",
"subspace",
"rank",
"tensor",
"norm",
"algorithm",
"sparsity",
"matrices"
],
[
"rl",
"reinforcement",
"robot",
"policy",
"reward",
"environments",
"environment",
"actions",
"agent",
"tasks"
],
[
"distance",
"clustering",
"kernel",
"graph",
"graphs",
"clusters",
"recovery",
"measure",
"between",
"structure"
],
[
"traffic",
"driving",
"demand",
"prediction",
"speed",
"spatial",
"vehicle",
"temporal",
"time",
"transportation"
],
[
"differential",
"systems",
"quantum",
"physics",
"equations",
"equation",
"parameters",
"informed",
"dynamics",
"physical"
],
[
"classification",
"eeg",
"features",
"health",
"svm",
"wearable",
"brain",
"signals",
"frequency",
"was"
],
[
"weather",
"day",
"forecasting",
"forecast",
"prediction",
"forecasts",
"power",
"forest",
"ensemble",
"different"
],
[
"vae",
"reconstruction",
"generative",
"object",
"gan",
"estimation",
"variational",
"latent",
"3d",
"distribution"
],
[
"text",
"research",
"researchers",
"software",
"ml",
"fact",
"social",
"health",
"media",
"open"
],
[
"missing",
"selection",
"causal",
"series",
"variables",
"inference",
"treatment",
"time",
"outcome",
"probabilistic"
],
[
"relu",
"networks",
"neural",
"pruning",
"layers",
"network",
"neurons",
"width",
"initialization",
"activation"
],
[
"images",
"image",
"video",
"visual",
"generative",
"object",
"gan",
"resolution",
"synthetic",
"cnn"
],
[
"algorithms",
"class",
"algorithm",
"tree",
"classes",
"clustering",
"classification",
"label",
"classifier",
"supervised"
],
[
"language",
"speech",
"audio",
"text",
"transformer",
"speaker",
"word",
"attention",
"languages",
"task"
],
[
"bandit",
"regret",
"agent",
"bound",
"arm",
"armed",
"sqrt",
"reward",
"online",
"bandits"
],
[
"effects",
"molecular",
"material",
"sequences",
"machine",
"protein",
"side",
"quantum",
"interactions",
"properties"
],
[
"gaussian",
"convergence",
"convex",
"stochastic",
"gradient",
"optimization",
"function",
"distributions",
"functions",
"variance"
],
[
"scores",
"fairness",
"face",
"population",
"bias",
"outcomes",
"fair",
"biases",
"groups",
"gender"
],
[
"recommendation",
"user",
"items",
"item",
"users",
"recommender",
"recommendations",
"filtering",
"collaborative",
"ranking"
],
[
"domain",
"feature",
"segmentation",
"supervised",
"image",
"deep",
"medical",
"target",
"unsupervised",
"classification"
],
[
"architecture",
"architectures",
"memory",
"nas",
"hardware",
"scaling",
"pruning",
"search",
"accelerators",
"neural"
],
[
"server",
"federated",
"communication",
"privacy",
"distributed",
"devices",
"local",
"private",
"decentralized",
"updates"
],
[
"adversarial",
"attacks",
"attack",
"against",
"robustness",
"defense",
"perturbations",
"robust",
"privacy",
"examples"
],
[
"embedding",
"gnns",
"graph",
"graphs",
"nodes",
"link",
"representation",
"embeddings",
"node",
"gnn"
],
[
"policy",
"reinforcement",
"games",
"agent",
"rl",
"policies",
"optimal",
"reward",
"value",
"control"
],
[
"19",
"image",
"ct",
"covid",
"patients",
"disease",
"were",
"images",
"cancer",
"segmentation"
]
] | 12.344938 | intfloat/e5-large-v2 | 0.863333 | -0.008441 | 0.146098 | 0.880327 |
ArXiv ML Papers | GMM | 42 | 40 | [
[
"family",
"distribution",
"pac",
"prior",
"weights",
"dropout",
"bayesian",
"uncertainty",
"posterior",
"predictive"
],
[
"explanation",
"explanations",
"counterfactual",
"interpretability",
"black",
"explainable",
"concepts",
"interpretable",
"ml",
"box"
],
[
"visual",
"driving",
"detection",
"object",
"images",
"cnn",
"video",
"frame",
"semantic",
"tracking"
],
[
"human",
"tensor",
"filter",
"machine",
"code",
"recurrent",
"train",
"al",
"project",
"was"
],
[
"rank",
"tensor",
"low",
"matrix",
"sparse",
"matrices",
"alternating",
"factorization",
"dictionary",
"recovery"
],
[
"reinforcement",
"agent",
"rl",
"robot",
"actions",
"policy",
"environment",
"reward",
"environments",
"tasks"
],
[
"space",
"distance",
"hierarchical",
"clustering",
"noise",
"nearest",
"measure",
"between",
"probabilistic",
"euclidean"
],
[
"prediction",
"demand",
"traffic",
"vehicle",
"transportation",
"modes",
"temporal",
"lstm",
"spatial",
"driving"
],
[
"equations",
"physics",
"quantum",
"dynamics",
"systems",
"differential",
"equation",
"physical",
"parameters",
"informed"
],
[
"series",
"eeg",
"signals",
"time",
"activity",
"features",
"wearable",
"frequency",
"recognition",
"sensors"
],
[
"day",
"forecasts",
"ensemble",
"forecasting",
"forecast",
"weather",
"predict",
"prediction",
"predicting",
"health"
],
[
"pose",
"object",
"reconstruction",
"distribution",
"estimation",
"test",
"inverse",
"calibration",
"uncertainty",
"robot"
],
[
"fact",
"social",
"health",
"researchers",
"ml",
"software",
"research",
"identity",
"questions",
"industry"
],
[
"graphical",
"series",
"variables",
"inference",
"selection",
"bayesian",
"causal",
"feature",
"likelihood",
"discovery"
],
[
"width",
"compositional",
"network",
"local",
"neurons",
"neural",
"networks",
"points",
"teacher",
"shape"
],
[
"image",
"gan",
"generative",
"images",
"gans",
"generator",
"music",
"quality",
"scene",
"generated"
],
[
"label",
"prediction",
"tree",
"classification",
"boosting",
"active",
"trees",
"algorithms",
"classifier",
"class"
],
[
"speaker",
"enhancement",
"audio",
"acoustic",
"end",
"speech",
"recognition",
"asr",
"transformer",
"signal"
],
[
"regret",
"bandit",
"bandits",
"arm",
"armed",
"ucb",
"reward",
"exploration",
"arms",
"sqrt"
],
[
"machine",
"molecular",
"sequences",
"ml",
"protein",
"property",
"properties",
"interactions",
"quantum",
"material"
],
[
"distribution",
"divergence",
"distributions",
"gaussian",
"processes",
"sampling",
"bayesian",
"bounds",
"fidelity",
"bound"
],
[
"fairness",
"fair",
"bias",
"face",
"gender",
"outcomes",
"algorithmic",
"groups",
"scores",
"population"
],
[
"social",
"label",
"users",
"items",
"user",
"collaborative",
"labels",
"features",
"preferences",
"channels"
],
[
"cell",
"net",
"images",
"medical",
"image",
"imaging",
"segmentation",
"dl",
"labeled",
"brain"
],
[
"memory",
"architectures",
"accelerators",
"hardware",
"search",
"nas",
"architecture",
"scaling",
"neural",
"bit"
],
[
"networks",
"initialization",
"relu",
"pruning",
"network",
"neural",
"layers",
"pooling",
"loss",
"deep"
],
[
"robustness",
"perturbations",
"attacks",
"against",
"defense",
"attack",
"examples",
"adversarial",
"robust",
"privacy"
],
[
"transfer",
"source",
"domain",
"target",
"feature",
"adaptation",
"domains",
"supervised",
"attention",
"unsupervised"
],
[
"rl",
"reinforcement",
"reward",
"policy",
"policies",
"agent",
"control",
"games",
"value",
"action"
],
[
"covid",
"image",
"images",
"dataset",
"patients",
"diagnosis",
"medical",
"convolutional",
"19",
"segmentation"
],
[
"privacy",
"communication",
"distributed",
"federated",
"server",
"private",
"local",
"devices",
"decentralized",
"updates"
],
[
"graph",
"graphs",
"node",
"representation",
"embedding",
"nodes",
"link",
"embeddings",
"structure",
"representations"
],
[
"vae",
"posterior",
"latent",
"distribution",
"variational",
"variables",
"inference",
"generative",
"approximate",
"density"
],
[
"security",
"attacks",
"detection",
"intrusion",
"service",
"traffic",
"iot",
"malicious",
"normal",
"attack"
],
[
"node",
"nodes",
"gnn",
"gnns",
"graph",
"spectral",
"graphs",
"passing",
"propagation",
"message"
],
[
"convergence",
"convex",
"stochastic",
"gradient",
"optimization",
"problems",
"descent",
"proximal",
"minimax",
"order"
],
[
"language",
"text",
"word",
"words",
"natural",
"translation",
"sentences",
"languages",
"semantic",
"attention"
],
[
"kernel",
"regression",
"clustering",
"linear",
"rank",
"manifold",
"subspace",
"subspaces",
"dimensionality",
"approximation"
],
[
"recommendation",
"user",
"item",
"recommender",
"items",
"sequence",
"recommendations",
"users",
"embedding",
"systems"
],
[
"regret",
"sqrt",
"online",
"bound",
"bounds",
"communication",
"optimal",
"sample",
"agent",
"policy"
]
] | 16.029501 | intfloat/e5-large-v2 | 0.825 | -0.02582 | 0.140531 | 0.883714 |
ArXiv ML Papers | GMM | 42 | 50 | [
[
"distributions",
"bayesian",
"posterior",
"weight",
"weights",
"networks",
"family",
"prior",
"variational",
"pac"
],
[
"explanation",
"black",
"concepts",
"counterfactual",
"explanations",
"interpretability",
"interpretable",
"box",
"predictions",
"logic"
],
[
"video",
"cnn",
"object",
"image",
"recognition",
"detection",
"channel",
"frame",
"proposal",
"images"
],
[
"them",
"al",
"assumptions",
"machine",
"train",
"tensor",
"human",
"library",
"topic",
"filter"
],
[
"tensor",
"matrix",
"rank",
"low",
"sparse",
"dictionary",
"matrices",
"alternating",
"recovery",
"factorization"
],
[
"games",
"reinforcement",
"rl",
"agent",
"agents",
"game",
"actions",
"atari",
"policy",
"environments"
],
[
"graph",
"distance",
"hierarchical",
"nearest",
"measure",
"space",
"noise",
"probabilistic",
"euclidean",
"communities"
],
[
"traffic",
"prediction",
"temporal",
"demand",
"lstm",
"recurrent",
"time",
"modes",
"service",
"attention"
],
[
"informed",
"dynamics",
"equation",
"systems",
"differential",
"physics",
"equations",
"nonlinear",
"physical",
"neural"
],
[
"eeg",
"temporal",
"frequency",
"features",
"series",
"time",
"signals",
"audio",
"forecasting",
"speech"
],
[
"weather",
"forecasts",
"prediction",
"developers",
"ensemble",
"software",
"day",
"forecast",
"forecasting",
"process"
],
[
"dropout",
"inputs",
"distribution",
"uncertainty",
"inverse",
"mnist",
"calibration",
"detection",
"test",
"dnns"
],
[
"social",
"ml",
"research",
"identity",
"health",
"fact",
"software",
"researchers",
"questions",
"about"
],
[
"bayesian",
"graphical",
"selection",
"inference",
"variables",
"causal",
"discovery",
"summary",
"likelihood",
"statistics"
],
[
"networks",
"number",
"infinite",
"points",
"compositional",
"width",
"classical",
"neural",
"kernels",
"functions"
],
[
"image",
"images",
"gans",
"gan",
"generative",
"generator",
"quality",
"music",
"generated",
"synthetic"
],
[
"clustering",
"cell",
"clusters",
"protein",
"order",
"expression",
"semi",
"feature",
"add",
"costs"
],
[
"speech",
"asr",
"acoustic",
"audio",
"speaker",
"recognition",
"end",
"enhancement",
"transformer",
"signal"
],
[
"user",
"strategy",
"query",
"reasoning",
"knowledge",
"algorithm",
"expert",
"active",
"queries",
"context"
],
[
"quantum",
"ml",
"sequences",
"machine",
"test",
"protein",
"tests",
"material",
"classifier",
"interactions"
],
[
"fidelity",
"processes",
"distributions",
"regression",
"gaussian",
"distribution",
"divergence",
"bounds",
"estimation",
"posterior"
],
[
"gender",
"bias",
"groups",
"fairness",
"fair",
"face",
"scores",
"population",
"outcomes",
"algorithmic"
],
[
"collaborative",
"labels",
"label",
"items",
"filtering",
"users",
"user",
"inference",
"preferences",
"profile"
],
[
"shape",
"imaging",
"medical",
"net",
"segmentation",
"image",
"images",
"brain",
"3d",
"mri"
],
[
"memory",
"architectures",
"nas",
"accelerators",
"hardware",
"neural",
"architecture",
"layer",
"search",
"precision"
],
[
"network",
"networks",
"relu",
"neural",
"layers",
"pooling",
"initialization",
"pruning",
"activation",
"bounds"
],
[
"adversarial",
"malware",
"influence",
"attacks",
"attack",
"poisoning",
"against",
"attacker",
"adversary",
"samples"
],
[
"supervised",
"source",
"domain",
"domains",
"feature",
"target",
"unsupervised",
"adaptation",
"transfer",
"metric"
],
[
"agents",
"agent",
"games",
"policy",
"reinforcement",
"action",
"gradient",
"stochastic",
"sum",
"value"
],
[
"covid",
"19",
"resolution",
"disease",
"xgboost",
"image",
"were",
"segmentation",
"convolutional",
"used"
],
[
"distributed",
"federated",
"updates",
"privacy",
"local",
"communication",
"devices",
"server",
"decentralized",
"centralized"
],
[
"representation",
"vertex",
"graph",
"embedding",
"nodes",
"link",
"node",
"structure",
"graphs",
"embeddings"
],
[
"variational",
"posterior",
"generative",
"distribution",
"variables",
"latent",
"inference",
"approximate",
"vae",
"coding"
],
[
"detection",
"iot",
"attacks",
"sensor",
"security",
"service",
"traffic",
"intrusion",
"detect",
"anomaly"
],
[
"gnn",
"gnns",
"spectral",
"node",
"graph",
"message",
"passing",
"graphs",
"nodes",
"propagation"
],
[
"stochastic",
"convex",
"descent",
"gradient",
"optimization",
"convergence",
"proximal",
"algorithm",
"problems",
"order"
],
[
"natural",
"text",
"language",
"word",
"translation",
"sentences",
"sentence",
"task",
"words",
"languages"
],
[
"regression",
"subspaces",
"clustering",
"manifold",
"approximation",
"subspace",
"rank",
"kernel",
"dimensionality",
"linear"
],
[
"recommender",
"user",
"items",
"item",
"recommendation",
"users",
"embedding",
"recommendations",
"systems",
"search"
],
[
"delta",
"optimal",
"bounds",
"sample",
"bound",
"mathcal",
"policy",
"regret",
"algorithms",
"lower"
],
[
"cancer",
"medical",
"patients",
"diseases",
"was",
"diagnosis",
"health",
"patient",
"disease",
"predict"
],
[
"bandit",
"bandits",
"regret",
"sqrt",
"arm",
"agents",
"agent",
"armed",
"ucb",
"reward"
],
[
"vehicles",
"autonomous",
"sensor",
"driving",
"vehicle",
"behavior",
"real",
"3d",
"depth",
"traffic"
],
[
"objects",
"robot",
"human",
"representations",
"object",
"visual",
"self",
"scene",
"pose",
"shot"
],
[
"reasoning",
"responses",
"visual",
"memory",
"representations",
"self",
"attention",
"recurrent",
"preference",
"interactions"
],
[
"cifar",
"student",
"group",
"distillation",
"knowledge",
"classifier",
"teacher",
"quantization",
"label",
"some"
],
[
"prediction",
"tree",
"classification",
"stream",
"ensemble",
"class",
"boosting",
"label",
"machine",
"classifiers"
],
[
"perturbations",
"attacks",
"adversarial",
"robustness",
"attack",
"robust",
"against",
"box",
"defense",
"examples"
],
[
"quantum",
"privacy",
"differential",
"private",
"sensitive",
"differentially",
"membership",
"sample",
"risks",
"guarantees"
],
[
"policy",
"reinforcement",
"rl",
"reward",
"control",
"policies",
"environment",
"environments",
"planning",
"goal"
]
] | 20.450282 | intfloat/e5-large-v2 | 0.784 | -0.033154 | 0.139364 | 0.880289 |
ArXiv ML Papers | KeyNMF | 42 | 10 | [
[
"analysis",
"statistical",
"how",
"distribution",
"information",
"methods",
"data",
"method",
"prediction",
"accuracy"
],
[
"optimization",
"algorithm",
"algorithms",
"optimal",
"methods",
"analysis",
"method",
"stochastic",
"gradient",
"performance"
],
[
"for",
"study",
"results",
"to",
"language",
"it",
"and",
"research",
"the",
"in"
],
[
"in",
"neural",
"graph",
"deep",
"convolutional",
"network",
"architecture",
"networks",
"accuracy",
"input"
],
[
"information",
"learn",
"prediction",
"reinforcement",
"representation",
"machine",
"deep",
"learning",
"representations",
"knowledge"
],
[
"methods",
"this",
"adversarial",
"number",
"method",
"for",
"to",
"problem",
"experiments",
"distribution"
],
[
"graph",
"features",
"where",
"graphs",
"information",
"text",
"to",
"from",
"knowledge",
"embeddings"
],
[
"accuracy",
"dataset",
"images",
"classification",
"image",
"datasets",
"features",
"detection",
"classifier",
"convolutional"
],
[
"experiments",
"techniques",
"examples",
"performance",
"training",
"adversarial",
"trained",
"task",
"robustness",
"generalization"
],
[
"datasets",
"model",
"models",
"generative",
"inference",
"estimation",
"modeling",
"bayesian",
"variational",
"latent"
]
] | 6.973114 | intfloat/e5-large-v2 | 0.76 | 0.019953 | 0.183277 | 0.662887 |
ArXiv ML Papers | KeyNMF | 42 | 20 | [
[
"distribution",
"data",
"statistical",
"how",
"probabilistic",
"computing",
"distributed",
"privacy",
"quality",
"predictive"
],
[
"optimal",
"sample",
"clustering",
"online",
"sparse",
"number",
"selection",
"algorithm",
"convergence",
"regret"
],
[
"to",
"and",
"for",
"in",
"it",
"language",
"the",
"study",
"that",
"research"
],
[
"convolutional",
"neural",
"networks",
"deep",
"network",
"architecture",
"input",
"in",
"accuracy",
"architectures"
],
[
"machine",
"learned",
"learning",
"reinforcement",
"learn",
"deep",
"prediction",
"policy",
"systems",
"environments"
],
[
"to",
"number",
"this",
"problem",
"distribution",
"for",
"generalization",
"example",
"error",
"experiments"
],
[
"trained",
"performance",
"experiments",
"training",
"task",
"generalization",
"speech",
"tasks",
"improve",
"techniques"
],
[
"classifiers",
"detection",
"classes",
"accuracy",
"class",
"classification",
"classifier",
"results",
"binary",
"label"
],
[
"similarity",
"features",
"matrix",
"feature",
"prediction",
"predict",
"convolutional",
"its",
"attention",
"speech"
],
[
"experiments",
"methods",
"method",
"clustering",
"regularization",
"techniques",
"outperforms",
"time",
"improve",
"process"
],
[
"modeling",
"model",
"inference",
"models",
"generative",
"prediction",
"estimation",
"machine",
"latent",
"experiments"
],
[
"text",
"to",
"from",
"detection",
"where",
"inference",
"and",
"how",
"trained",
"samples"
],
[
"research",
"applications",
"human",
"datasets",
"labels",
"dataset",
"evaluation",
"label",
"regularization",
"bias"
],
[
"where",
"information",
"communication",
"representations",
"semantic",
"estimation",
"probability",
"privacy",
"feature",
"representation"
],
[
"how",
"applications",
"techniques",
"results",
"algorithms",
"efficiency",
"where",
"problem",
"systems",
"performance"
],
[
"images",
"image",
"convolutional",
"deep",
"generative",
"segmentation",
"imaging",
"representations",
"computer",
"trained"
],
[
"examples",
"robust",
"attack",
"adversarial",
"attacks",
"robustness",
"detection",
"defense",
"security",
"perturbations"
],
[
"sample",
"analysis",
"study",
"theoretical",
"applications",
"complexity",
"research",
"statistical",
"results",
"researchers"
],
[
"bayesian",
"optimal",
"gaussian",
"gradient",
"stochastic",
"optimization",
"performance",
"approximation",
"inference",
"sampling"
],
[
"network",
"graphs",
"nodes",
"graph",
"networks",
"node",
"representations",
"prediction",
"knowledge",
"representation"
]
] | 7.027881 | intfloat/e5-large-v2 | 0.7 | -0.024959 | 0.158717 | 0.724167 |
ArXiv ML Papers | KeyNMF | 42 | 30 | [
[
"quality",
"statistical",
"distributed",
"data",
"computing",
"privacy",
"computation",
"techniques",
"technology",
"benchmarks"
],
[
"applications",
"techniques",
"matrix",
"results",
"algorithms",
"efficiency",
"computational",
"where",
"problem",
"probability"
],
[
"language",
"in",
"time",
"detection",
"human",
"it",
"attention",
"its",
"memory",
"outperforms"
],
[
"recurrent",
"neural",
"convolutional",
"networks",
"input",
"architecture",
"deep",
"architectures",
"these",
"dnn"
],
[
"machine",
"learn",
"learns",
"reinforcement",
"knowledge",
"learned",
"learning",
"deep",
"policy",
"environments"
],
[
"number",
"problem",
"theoretical",
"generalization",
"error",
"example",
"this",
"where",
"its",
"loss"
],
[
"speech",
"transfer",
"task",
"trained",
"generalization",
"tasks",
"training",
"techniques",
"examples",
"target"
],
[
"label",
"classification",
"detection",
"classes",
"class",
"classifiers",
"classifier",
"binary",
"labels",
"feature"
],
[
"the",
"of",
"that",
"it",
"to",
"results",
"or",
"for",
"and",
"language"
],
[
"process",
"method",
"regularization",
"outperforms",
"clustering",
"methods",
"time",
"techniques",
"improve",
"for"
],
[
"machine",
"generative",
"modeling",
"these",
"estimation",
"model",
"models",
"parameters",
"latent",
"design"
],
[
"text",
"where",
"detection",
"knowledge",
"samples",
"from",
"trained",
"software",
"code",
"github"
],
[
"applications",
"datasets",
"dataset",
"research",
"labels",
"human",
"computer",
"evaluation",
"label",
"bias"
],
[
"information",
"where",
"semantic",
"feature",
"communication",
"representations",
"privacy",
"tasks",
"share",
"compression"
],
[
"predict",
"studies",
"study",
"research",
"disease",
"health",
"researchers",
"machine",
"clinical",
"sample"
],
[
"image",
"images",
"convolutional",
"deep",
"generative",
"imaging",
"segmentation",
"computer",
"representations",
"medical"
],
[
"adversarial",
"examples",
"attack",
"attacks",
"robustness",
"robust",
"detection",
"defense",
"security",
"perturbations"
],
[
"statistical",
"results",
"computer",
"sample",
"theoretical",
"analysis",
"complexity",
"researchers",
"applications",
"characteristics"
],
[
"distributions",
"estimation",
"probability",
"generalization",
"distribution",
"sample",
"samples",
"density",
"sampling",
"likelihood"
],
[
"graph",
"graphs",
"representations",
"representation",
"nodes",
"node",
"networks",
"embedding",
"embeddings",
"link"
],
[
"gradient",
"stochastic",
"convergence",
"solving",
"method",
"optimization",
"convex",
"optimal",
"problem",
"solutions"
],
[
"trained",
"accuracy",
"improve",
"results",
"dataset",
"computational",
"result",
"efficiency",
"deep",
"devices"
],
[
"features",
"feature",
"attention",
"speech",
"similarity",
"its",
"convolutional",
"matrix",
"relevant",
"structure"
],
[
"optimal",
"computational",
"memory",
"parameters",
"systems",
"evaluation",
"efficiency",
"robust",
"performance",
"metrics"
],
[
"optimal",
"number",
"convergence",
"sparse",
"algorithm",
"selection",
"regret",
"online",
"sample",
"clustering"
],
[
"predictions",
"predict",
"predictive",
"prediction",
"predicting",
"sequence",
"predicted",
"link",
"outperforms",
"future"
],
[
"variational",
"inference",
"gaussian",
"stochastic",
"bayesian",
"computational",
"latent",
"processes",
"approximate",
"distributions"
],
[
"memory",
"communication",
"number",
"distributed",
"network",
"networks",
"detection",
"architecture",
"knowledge",
"nodes"
],
[
"solutions",
"search",
"knowledge",
"experiments",
"quality",
"reinforcement",
"generate",
"reward",
"improve",
"to"
],
[
"how",
"learn",
"dynamics",
"representations",
"time",
"tasks",
"applications",
"systems",
"task",
"human"
]
] | 6.735505 | intfloat/e5-large-v2 | 0.636667 | -0.029963 | 0.167181 | 0.765255 |
ArXiv ML Papers | KeyNMF | 42 | 40 | [
[
"statistical",
"quality",
"distributed",
"computing",
"data",
"privacy",
"technology",
"techniques",
"benchmarks",
"predictive"
],
[
"statistical",
"computational",
"matrix",
"complexity",
"memory",
"computation",
"efficiency",
"sparse",
"approximation",
"processing"
],
[
"human",
"in",
"time",
"it",
"its",
"speech",
"outperforms",
"new",
"audio",
"where"
],
[
"these",
"it",
"solutions",
"deep",
"dnn",
"here",
"dl",
"applied",
"dnns",
"autoencoder"
],
[
"learning",
"active",
"implementation",
"systems",
"machine",
"ml",
"learns",
"online",
"representation",
"science"
],
[
"its",
"this",
"problem",
"error",
"generalization",
"number",
"example",
"loss",
"theoretical",
"where"
],
[
"training",
"test",
"federated",
"loss",
"input",
"improve",
"examples",
"generalization",
"techniques",
"annotated"
],
[
"classifier",
"class",
"classifiers",
"classification",
"classes",
"label",
"labels",
"binary",
"feature",
"classify"
],
[
"representations",
"transfer",
"knowledge",
"tasks",
"trained",
"language",
"task",
"target",
"speech",
"representation"
],
[
"process",
"method",
"clustering",
"problem",
"number",
"technique",
"clusters",
"time",
"regularization",
"outperforms"
],
[
"parameters",
"these",
"models",
"estimation",
"machine",
"modeling",
"generative",
"model",
"learn",
"privacy"
],
[
"outperforms",
"samples",
"from",
"where",
"text",
"software",
"knowledge",
"users",
"online",
"identifying"
],
[
"dataset",
"bias",
"learn",
"datasets",
"labels",
"human",
"label",
"regularization",
"research",
"evaluation"
],
[
"optimal",
"problem",
"complexity",
"regret",
"linear",
"stochastic",
"number",
"approximation",
"gradient",
"sample"
],
[
"matrix",
"techniques",
"results",
"search",
"where",
"problem",
"clustering",
"algorithms",
"probability",
"research"
],
[
"image",
"images",
"convolutional",
"segmentation",
"computer",
"imaging",
"medical",
"generative",
"vision",
"representations"
],
[
"adversarial",
"attack",
"examples",
"attacks",
"robust",
"robustness",
"defense",
"generative",
"perturbations",
"security"
],
[
"complexity",
"statistical",
"analysis",
"sample",
"computer",
"results",
"theoretical",
"researchers",
"sentiment",
"characteristics"
],
[
"distribution",
"distributions",
"probability",
"samples",
"sampling",
"estimation",
"density",
"likelihood",
"generalization",
"generative"
],
[
"representation",
"node",
"embeddings",
"nodes",
"graphs",
"graph",
"embedding",
"representations",
"link",
"knowledge"
],
[
"solving",
"stochastic",
"convergence",
"optimization",
"convex",
"gradient",
"solutions",
"techniques",
"functions",
"space"
],
[
"dataset",
"result",
"accuracy",
"improve",
"results",
"improving",
"speed",
"imagenet",
"output",
"developed"
],
[
"its",
"feature",
"attention",
"convolutional",
"features",
"speech",
"similarity",
"matrix",
"relevant",
"structure"
],
[
"memory",
"performance",
"parameters",
"robust",
"systems",
"evaluation",
"metrics",
"results",
"lstm",
"efficiency"
],
[
"selection",
"algorithm",
"convergence",
"matrix",
"sample",
"regret",
"clustering",
"sparse",
"efficiency",
"adaptive"
],
[
"predictive",
"predict",
"prediction",
"predictions",
"predicting",
"predicted",
"link",
"sequence",
"outperforms",
"regression"
],
[
"semantic",
"information",
"representations",
"privacy",
"probability",
"feature",
"communication",
"share",
"local",
"where"
],
[
"architectures",
"number",
"knowledge",
"network",
"architecture",
"communication",
"distributed",
"nodes",
"convergence",
"federated"
],
[
"experiments",
"search",
"knowledge",
"solutions",
"quality",
"improve",
"extensive",
"testing",
"sparse",
"target"
],
[
"to",
"it",
"text",
"that",
"task",
"the",
"for",
"language",
"attention",
"of"
],
[
"networks",
"parameters",
"computation",
"these",
"theory",
"memory",
"theoretical",
"relu",
"activations",
"layers"
],
[
"policy",
"reinforcement",
"learn",
"environments",
"learned",
"agents",
"control",
"policies",
"rl",
"tasks"
],
[
"methods",
"improve",
"empirical",
"techniques",
"sampling",
"numerical",
"distance",
"regularization",
"samples",
"outperforms"
],
[
"computer",
"applications",
"computing",
"processing",
"computation",
"architecture",
"research",
"devices",
"users",
"user"
],
[
"bayesian",
"approximate",
"distributions",
"variables",
"variational",
"latent",
"gaussian",
"inference",
"stochastic",
"processes"
],
[
"dataset",
"object",
"detect",
"detecting",
"detection",
"it",
"anomaly",
"detected",
"attacks",
"detector"
],
[
"clinical",
"predict",
"machine",
"researchers",
"studies",
"research",
"health",
"study",
"disease",
"methodology"
],
[
"and",
"results",
"linear",
"for",
"the",
"of",
"gradient",
"or",
"regression",
"loss"
],
[
"architectures",
"architecture",
"recurrent",
"function",
"convolutional",
"neural",
"input",
"techniques",
"design",
"physics"
],
[
"how",
"systems",
"concepts",
"interactions",
"time",
"dynamics",
"future",
"example",
"fairness",
"brain"
]
] | 8.592247 | intfloat/e5-large-v2 | 0.6225 | -0.065104 | 0.16314 | 0.790585 |
ArXiv ML Papers | KeyNMF | 42 | 50 | [
[
"computing",
"distributed",
"privacy",
"statistical",
"data",
"quality",
"technology",
"predictive",
"simulation",
"outcome"
],
[
"learning",
"systems",
"ml",
"machine",
"active",
"learns",
"science",
"researchers",
"studies",
"generalization"
],
[
"and",
"the",
"or",
"for",
"of",
"linear",
"that",
"also",
"loss",
"new"
],
[
"architectures",
"recurrent",
"physics",
"functions",
"input",
"techniques",
"architecture",
"function",
"neural",
"neurons"
],
[
"dnn",
"here",
"deep",
"these",
"solutions",
"dnns",
"dl",
"applied",
"autoencoder",
"medical"
],
[
"this",
"problem",
"example",
"theoretical",
"loss",
"generalization",
"where",
"context",
"error",
"these"
],
[
"federated",
"training",
"generalization",
"improve",
"task",
"techniques",
"loss",
"examples",
"annotated",
"computation"
],
[
"classification",
"classifier",
"class",
"feature",
"classifiers",
"classes",
"labels",
"label",
"binary",
"classify"
],
[
"for",
"to",
"of",
"that",
"the",
"or",
"on",
"online",
"real",
"attention"
],
[
"clustering",
"method",
"process",
"problem",
"technique",
"clusters",
"outperforms",
"regularization",
"classifier",
"processes"
],
[
"generative",
"estimation",
"model",
"modeling",
"machine",
"models",
"these",
"parameters",
"engineering",
"privacy"
],
[
"samples",
"github",
"where",
"text",
"from",
"outperforms",
"software",
"users",
"identifying",
"code"
],
[
"overfitting",
"computer",
"regularization",
"human",
"labels",
"bias",
"evaluation",
"label",
"real",
"datasets"
],
[
"privacy",
"probability",
"where",
"information",
"communication",
"feature",
"share",
"local",
"semantic",
"compression"
],
[
"algorithms",
"where",
"problem",
"implementation",
"search",
"clustering",
"research",
"matrix",
"comparison",
"probability"
],
[
"image",
"computer",
"gan",
"segmentation",
"art",
"imaging",
"images",
"generative",
"medical",
"real"
],
[
"security",
"attack",
"perturbations",
"generative",
"robust",
"robustness",
"defense",
"adversarial",
"examples",
"attacks"
],
[
"statistical",
"complexity",
"computer",
"sample",
"theoretical",
"analysis",
"researchers",
"sentiment",
"structures",
"technologies"
],
[
"distribution",
"probability",
"estimation",
"density",
"distributions",
"samples",
"sampling",
"generative",
"generalization",
"likelihood"
],
[
"graphs",
"link",
"node",
"graph",
"embedding",
"matrix",
"clustering",
"nodes",
"embeddings",
"structure"
],
[
"functions",
"convex",
"solutions",
"optimization",
"space",
"function",
"objective",
"solved",
"search",
"techniques"
],
[
"result",
"improve",
"accuracy",
"imagenet",
"speed",
"devices",
"improving",
"output",
"improvement",
"identification"
],
[
"speech",
"its",
"similarity",
"attention",
"relevant",
"feature",
"features",
"interpretability",
"matrix",
"correlations"
],
[
"memory",
"parameters",
"systems",
"robust",
"performance",
"evaluation",
"lstm",
"error",
"metrics",
"estimate"
],
[
"regret",
"selection",
"algorithm",
"convergence",
"matrix",
"online",
"adaptive",
"efficiency",
"sample",
"clustering"
],
[
"sequence",
"predictions",
"predict",
"predictive",
"prediction",
"predicting",
"predicted",
"link",
"outperforms",
"future"
],
[
"variational",
"latent",
"inference",
"gaussian",
"distributions",
"bayesian",
"processes",
"approximate",
"variables",
"probabilistic"
],
[
"nodes",
"distributed",
"architectures",
"federated",
"architecture",
"node",
"convergence",
"communication",
"network",
"devices"
],
[
"extensive",
"quality",
"experiments",
"search",
"solutions",
"techniques",
"improve",
"testing",
"sequences",
"variables"
],
[
"objects",
"github",
"object",
"visual",
"contrastive",
"representation",
"embeddings",
"embedding",
"self",
"representations"
],
[
"computation",
"these",
"networks",
"parameters",
"theoretical",
"relu",
"memory",
"activations",
"depth",
"theory"
],
[
"solution",
"optimal",
"problem",
"regret",
"sample",
"approximation",
"complexity",
"linear",
"online",
"processes"
],
[
"methods",
"improve",
"empirical",
"samples",
"sampling",
"regularization",
"outperforms",
"implementation",
"distance",
"evaluation"
],
[
"time",
"interactions",
"future",
"concepts",
"systems",
"how",
"dynamics",
"example",
"brain",
"fairness"
],
[
"computational",
"efficiency",
"statistical",
"memory",
"complexity",
"computation",
"matrix",
"processing",
"approximation",
"theoretical"
],
[
"detected",
"detect",
"anomaly",
"detecting",
"attacks",
"video",
"detection",
"detector",
"object",
"anomalies"
],
[
"research",
"science",
"clinical",
"researchers",
"health",
"studies",
"disease",
"study",
"predict",
"methodology"
],
[
"policy",
"reinforcement",
"agents",
"environments",
"rl",
"action",
"policies",
"control",
"reward",
"efficiency"
],
[
"human",
"where",
"new",
"humans",
"time",
"memory",
"in",
"outperforms",
"vector",
"model"
],
[
"convolution",
"convolutional",
"architecture",
"cnn",
"processing",
"attention",
"segmentation",
"outperforms",
"research",
"gpu"
],
[
"stochastic",
"convex",
"variance",
"solving",
"convergence",
"gradient",
"gradients",
"equations",
"dynamics",
"approximate"
],
[
"sample",
"theoretical",
"results",
"testing",
"experimental",
"clustering",
"techniques",
"complexity",
"result",
"regression"
],
[
"text",
"task",
"word",
"tasks",
"language",
"words",
"the",
"languages",
"speech",
"sentence"
],
[
"research",
"dataset",
"quality",
"lstm",
"class",
"2019",
"source",
"comparison",
"task",
"mnist"
],
[
"learn",
"learned",
"human",
"learns",
"tasks",
"humans",
"visual",
"dynamic",
"dynamics",
"task"
],
[
"that",
"evaluation",
"machine",
"software",
"its",
"quality",
"process",
"problem",
"test",
"it"
],
[
"computation",
"computing",
"applications",
"processing",
"users",
"devices",
"compression",
"computer",
"research",
"user"
],
[
"knowledge",
"domain",
"transfer",
"online",
"target",
"adaptation",
"domains",
"reasoning",
"distillation",
"source"
],
[
"number",
"sparse",
"numerical",
"linear",
"regression",
"examples",
"error",
"gaussian",
"theory",
"parameters"
],
[
"improvement",
"example",
"task",
"code",
"trained",
"transfer",
"tasks",
"github",
"target",
"memory"
]
] | 9.060088 | intfloat/e5-large-v2 | 0.6 | -0.092257 | 0.169728 | 0.810261 |
ArXiv ML Papers | FASTopic | 42 | 10 | [
[
"neural",
"network",
"networks",
"adversarial",
"graph",
"image",
"input",
"loss",
"structure",
"scale"
],
[
"attention",
"language",
"modeling",
"source",
"natural",
"speech",
"sequence",
"generation",
"visual",
"semantic"
],
[
"algorithms",
"function",
"problems",
"problem",
"algorithm",
"optimization",
"non",
"gradient",
"number",
"where"
],
[
"family",
"approximate",
"nodes",
"probabilistic",
"product",
"weighted",
"missing",
"scalable",
"private",
"exact"
],
[
"tree",
"per",
"steps",
"popular",
"fast",
"faster",
"batch",
"forecasting",
"wide",
"sequential"
],
[
"what",
"dynamic",
"simulation",
"states",
"programming",
"concepts",
"remains",
"planning",
"game",
"enable"
],
[
"graphs",
"components",
"tensor",
"node",
"flow",
"vector",
"structured",
"tool",
"hierarchical",
"interactions"
],
[
"testing",
"community",
"importance",
"binary",
"validation",
"compare",
"identifying",
"collection",
"would",
"correct"
],
[
"task",
"prediction",
"features",
"different",
"deep",
"tasks",
"models",
"machine",
"classification",
"datasets"
],
[
"developed",
"cnn",
"github",
"architectures",
"temporal",
"predict",
"https",
"com",
"layers",
"processing"
]
] | 86.729294 | intfloat/e5-large-v2 | 1 | -0.183123 | 0.128093 | 0.838472 |
ArXiv ML Papers | FASTopic | 42 | 20 | [
[
"superior",
"informative",
"either",
"dependencies",
"handle",
"typical",
"advantage",
"core",
"pair",
"interpretation"
],
[
"rnn",
"recurrent",
"words",
"style",
"languages",
"encoding",
"topic",
"named",
"decoding",
"transformers"
],
[
"iteration",
"assumptions",
"equations",
"near",
"special",
"constant",
"approximations",
"stability",
"strongly",
"sum"
],
[
"how",
"machine",
"systems",
"process",
"system",
"detection",
"human",
"real",
"use",
"dataset"
],
[
"domain",
"knowledge",
"tasks",
"task",
"adversarial",
"level",
"attention",
"representations",
"research",
"language"
],
[
"graph",
"feature",
"samples",
"structure",
"inference",
"clustering",
"matrix",
"space",
"estimation",
"regression"
],
[
"start",
"affect",
"review",
"growing",
"focuses",
"past",
"every",
"needs",
"developing",
"done"
],
[
"autonomous",
"path",
"dynamic",
"simulated",
"simulation",
"physics",
"initial",
"forward",
"update",
"formulate"
],
[
"physical",
"achieving",
"hardware",
"implementation",
"modern",
"scaling",
"capable",
"mobile",
"frameworks",
"resources"
],
[
"label",
"related",
"so",
"form",
"provides",
"vector",
"develop",
"series",
"some",
"support"
],
[
"family",
"probabilistic",
"suitable",
"limit",
"exact",
"discrete",
"propagation",
"estimating",
"version",
"flows"
],
[
"rank",
"tensor",
"phase",
"weighted",
"product",
"missing",
"spectral",
"points",
"sparsity",
"vectors"
],
[
"signals",
"signal",
"maps",
"forecasting",
"demand",
"area",
"regions",
"events",
"short",
"event"
],
[
"called",
"expensive",
"leverage",
"hidden",
"competitive",
"relative",
"manner",
"leading",
"preserving",
"explicit"
],
[
"neural",
"networks",
"deep",
"network",
"image",
"images",
"loss",
"training",
"scale",
"convolutional"
],
[
"function",
"optimization",
"gradient",
"optimal",
"stochastic",
"functions",
"complexity",
"policy",
"sample",
"convergence"
],
[
"instead",
"little",
"another",
"soft",
"fidelity",
"discover",
"capacity",
"called",
"auxiliary",
"pose"
],
[
"conventional",
"precision",
"applied",
"speed",
"crucial",
"utilizing",
"co",
"proposes",
"namely",
"performing"
],
[
"parallel",
"programming",
"tree",
"al",
"et",
"subset",
"reduces",
"utility",
"algorithmic",
"true"
],
[
"in",
"of",
"the",
"to",
"and",
"we",
"is",
"on",
"that",
"for"
]
] | 1,786.851234 | intfloat/e5-large-v2 | 0.995 | -0.234208 | 0.128766 | 0.856957 |
ArXiv ML Papers | FASTopic | 42 | 30 | [
[
"tree",
"selected",
"utility",
"processes",
"sequential",
"costs",
"per",
"algorithmic",
"simulations",
"fair"
],
[
"expensive",
"faster",
"fast",
"reduced",
"parallel",
"tuning",
"comparable",
"reduces",
"solve",
"allows"
],
[
"jointly",
"diversity",
"sequences",
"autoencoders",
"competitive",
"superior",
"hidden",
"benchmarks",
"consistently",
"joint"
],
[
"demand",
"forecasting",
"community",
"article",
"methodology",
"forecast",
"sources",
"survey",
"outcomes",
"life"
],
[
"environments",
"games",
"exploration",
"simulation",
"policies",
"reward",
"action",
"actions",
"meta",
"bandit"
],
[
"neural",
"networks",
"network",
"deep",
"graph",
"end",
"convolutional",
"architecture",
"memory",
"layer"
],
[
"weight",
"edge",
"speed",
"batch",
"mnist",
"computation",
"flow",
"physics",
"compression",
"geometric"
],
[
"computing",
"distributed",
"probabilistic",
"al",
"et",
"queries",
"exact",
"entropy",
"subset",
"estimating"
],
[
"function",
"sample",
"linear",
"gradient",
"distribution",
"optimization",
"optimal",
"local",
"functions",
"stochastic"
],
[
"algorithms",
"problem",
"algorithm",
"problems",
"analysis",
"order",
"noise",
"given",
"any",
"number"
],
[
"style",
"audio",
"recurrent",
"module",
"words",
"event",
"reasoning",
"transformer",
"translation",
"convolution"
],
[
"complete",
"automatically",
"review",
"quantitative",
"open",
"insights",
"physical",
"building",
"progress",
"evaluating"
],
[
"detection",
"samples",
"robust",
"attacks",
"adversarial",
"prediction",
"examples",
"classifier",
"series",
"robustness"
],
[
"comprehensive",
"made",
"crucial",
"open",
"namely",
"levels",
"having",
"side",
"numerous",
"physical"
],
[
"concepts",
"topic",
"interpretability",
"relational",
"explanations",
"aspects",
"biases",
"agnostic",
"discover",
"explain"
],
[
"modern",
"needs",
"execution",
"easily",
"verification",
"affect",
"every",
"frameworks",
"requirements",
"deployment"
],
[
"ml",
"was",
"human",
"were",
"system",
"machine",
"developed",
"research",
"potential",
"behavior"
],
[
"strongly",
"sum",
"approximate",
"approximations",
"log",
"iteration",
"special",
"markov",
"monte",
"divergence"
],
[
"image",
"features",
"feature",
"supervised",
"images",
"latent",
"representation",
"clustering",
"speech",
"object"
],
[
"task",
"tasks",
"trained",
"source",
"language",
"models",
"text",
"modeling",
"domain",
"generative"
],
[
"design",
"reinforcement",
"policy",
"user",
"framework",
"cost",
"knowledge",
"decision",
"online",
"control"
],
[
"extend",
"performs",
"dynamical",
"implement",
"thereby",
"carefully",
"perturbed",
"applicable",
"considered",
"far"
],
[
"usage",
"mobile",
"collected",
"boosting",
"sensor",
"events",
"monitoring",
"cloud",
"day",
"people"
],
[
"after",
"typical",
"particularly",
"take",
"entire",
"allowing",
"idea",
"comparing",
"quickly",
"proposes"
],
[
"spectral",
"tensor",
"vectors",
"product",
"property",
"phase",
"decomposition",
"equations",
"matrices",
"measurements"
],
[
"link",
"suitable",
"cluster",
"weighted",
"matching",
"scalable",
"schemes",
"theoretically",
"compute",
"construct"
],
[
"precision",
"resolution",
"regions",
"hardware",
"region",
"pipeline",
"final",
"100",
"frequency",
"device"
],
[
"term",
"presents",
"relationship",
"protein",
"change",
"changes",
"mapping",
"part",
"represent",
"conducted"
],
[
"of",
"we",
"is",
"the",
"in",
"that",
"to",
"and",
"on",
"for"
],
[
"calibration",
"negative",
"outlier",
"imbalanced",
"positive",
"consistent",
"weak",
"binary",
"varying",
"calibrated"
]
] | 1,947.61711 | intfloat/e5-large-v2 | 0.993333 | -0.247534 | 0.135523 | 0.873651 |
ArXiv ML Papers | FASTopic | 42 | 40 | [
[
"understanding",
"speech",
"signals",
"text",
"sequence",
"audio",
"language",
"recurrent",
"word",
"sequences"
],
[
"behaviors",
"items",
"states",
"decisions",
"motion",
"autonomous",
"experience",
"feasible",
"gain",
"own"
],
[
"end",
"images",
"memory",
"convolutional",
"architecture",
"pre",
"recognition",
"segmentation",
"image",
"cnn"
],
[
"vulnerability",
"post",
"trigger",
"crafted",
"white",
"defend",
"capable",
"generates",
"indicate",
"vulnerable"
],
[
"reducing",
"physics",
"capacity",
"fidelity",
"coding",
"operations",
"vae",
"forward",
"quantization",
"operation"
],
[
"implicit",
"leveraging",
"margin",
"proposes",
"augmented",
"inspired",
"dependencies",
"explicit",
"represented",
"avoid"
],
[
"optimization",
"gradient",
"solution",
"convergence",
"communication",
"efficient",
"parameter",
"popular",
"stochastic",
"convex"
],
[
"fields",
"filter",
"compression",
"propagation",
"sensitivity",
"flows",
"computed",
"diffusion",
"invariant",
"filters"
],
[
"selection",
"distribution",
"prior",
"inference",
"bayesian",
"distributions",
"variational",
"search",
"variables",
"space"
],
[
"events",
"boosting",
"areas",
"short",
"forecast",
"usage",
"five",
"six",
"sensors",
"location"
],
[
"combine",
"associated",
"assess",
"tools",
"tool",
"sources",
"entire",
"presented",
"incorporated",
"reliability"
],
[
"basic",
"version",
"operators",
"randomly",
"sufficiently",
"sizes",
"takes",
"solved",
"nn",
"smaller"
],
[
"allows",
"formulate",
"response",
"steps",
"stage",
"differentiable",
"beyond",
"heuristic",
"designing",
"maximize"
],
[
"together",
"corresponding",
"instances",
"sufficient",
"sampled",
"stream",
"streaming",
"ratio",
"established",
"interesting"
],
[
"library",
"body",
"requirements",
"implementation",
"surface",
"gpu",
"offers",
"magnetic",
"digital",
"technical"
],
[
"rather",
"agnostic",
"mechanisms",
"evidence",
"enable",
"represent",
"explore",
"interpretability",
"concepts",
"building"
],
[
"costs",
"residual",
"magnitude",
"flexible",
"parallel",
"speed",
"additionally",
"reduces",
"evaluations",
"achieving"
],
[
"source",
"graph",
"attention",
"representation",
"representations",
"object",
"unsupervised",
"graphs",
"visual",
"semantic"
],
[
"development",
"help",
"community",
"researchers",
"questions",
"software",
"engineering",
"intelligence",
"public",
"humans"
],
[
"matrix",
"dimensional",
"clustering",
"metric",
"regression",
"point",
"kernel",
"low",
"vector",
"points"
],
[
"deep",
"network",
"neural",
"training",
"networks",
"layer",
"adversarial",
"attacks",
"input",
"robustness"
],
[
"showed",
"assessment",
"day",
"early",
"2019",
"forest",
"report",
"database",
"svm",
"20"
],
[
"mapping",
"change",
"tracking",
"experiment",
"channels",
"employed",
"separate",
"variations",
"pose",
"verification"
],
[
"system",
"systems",
"human",
"about",
"current",
"user",
"bias",
"world",
"machine",
"making"
],
[
"identifying",
"review",
"scientific",
"science",
"build",
"topic",
"growing",
"aspects",
"reference",
"complete"
],
[
"temporal",
"power",
"uses",
"was",
"series",
"driven",
"developed",
"were",
"term",
"signal"
],
[
"take",
"next",
"deployed",
"auxiliary",
"potentially",
"tackle",
"mitigate",
"observe",
"aims",
"another"
],
[
"deterministic",
"left",
"exponential",
"delay",
"theoretic",
"linearly",
"em",
"randomized",
"proof",
"mcmc"
],
[
"rate",
"generalization",
"bounds",
"estimation",
"bound",
"robust",
"error",
"noise",
"sample",
"uncertainty"
],
[
"outcomes",
"appropriate",
"alternative",
"correct",
"concept",
"discovery",
"notion",
"algorithmic",
"relational",
"biases"
],
[
"grained",
"part",
"generating",
"seen",
"easily",
"introducing",
"combined",
"typical",
"according",
"taking"
],
[
"solve",
"reinforcement",
"agents",
"agent",
"policy",
"control",
"rl",
"value",
"continuous",
"reward"
],
[
"al",
"et",
"approximately",
"trees",
"tree",
"total",
"subset",
"programming",
"yield",
"contribution"
],
[
"propose",
"state",
"multi",
"tasks",
"learning",
"knowledge",
"learn",
"domain",
"framework",
"design"
],
[
"positive",
"negative",
"vectors",
"pair",
"outlier",
"core",
"filtering",
"construction",
"correlations",
"deal"
],
[
"the",
"in",
"of",
"and",
"to",
"we",
"is",
"that",
"for",
"on"
],
[
"variation",
"estimating",
"statistics",
"decomposition",
"basis",
"graphical",
"mixture",
"correlated",
"treatment",
"neighbor"
],
[
"prediction",
"dataset",
"classification",
"datasets",
"detection",
"features",
"accuracy",
"test",
"feature",
"classifier"
],
[
"use",
"task",
"research",
"complex",
"models",
"specific",
"modeling",
"quality",
"natural",
"generation"
],
[
"algorithm",
"problem",
"problems",
"algorithms",
"optimal",
"number",
"where",
"given",
"any",
"consider"
]
] | 1,903.247297 | intfloat/e5-large-v2 | 1 | -0.243696 | 0.130374 | 0.877358 |
ArXiv ML Papers | FASTopic | 42 | 50 | [
[
"network",
"neural",
"deep",
"networks",
"generalization",
"architecture",
"training",
"layer",
"loss",
"input"
],
[
"classified",
"performances",
"change",
"patch",
"distinct",
"capabilities",
"generalizability",
"presents",
"mode",
"separate"
],
[
"tasks",
"task",
"knowledge",
"end",
"multi",
"language",
"domain",
"level",
"attention",
"outperforms"
],
[
"interpretable",
"concepts",
"mechanisms",
"represent",
"agnostic",
"generating",
"decisions",
"interpretability",
"meaningful",
"explanations"
],
[
"assessment",
"svm",
"day",
"forest",
"people",
"showed",
"2019",
"eeg",
"reports",
"healthcare"
],
[
"implicit",
"physics",
"soft",
"explicitly",
"joint",
"hidden",
"explicit",
"represented",
"dependencies",
"yields"
],
[
"game",
"motion",
"autonomous",
"vae",
"environments",
"planning",
"experience",
"trajectories",
"imitation",
"own"
],
[
"mostly",
"mode",
"take",
"always",
"analyses",
"white",
"representative",
"coverage",
"plays",
"conclude"
],
[
"feature",
"classification",
"ensemble",
"features",
"classes",
"flow",
"because",
"spatial",
"group",
"means"
],
[
"optimized",
"resources",
"hardware",
"utilizing",
"physical",
"implementation",
"frequency",
"frameworks",
"tracking",
"demonstrates"
],
[
"vectors",
"basis",
"negative",
"propagation",
"fields",
"arbitrary",
"pair",
"filters",
"geometry",
"edges"
],
[
"learning",
"more",
"not",
"but",
"one",
"work",
"when",
"their",
"only",
"information"
],
[
"presented",
"leverages",
"methodology",
"attracted",
"reasonable",
"combined",
"produced",
"cause",
"performing",
"addresses"
],
[
"rnn",
"style",
"question",
"translation",
"languages",
"words",
"decoder",
"reasoning",
"topic",
"encoding"
],
[
"forecasting",
"lstm",
"demand",
"short",
"life",
"sensors",
"location",
"sensor",
"forecast",
"intelligent"
],
[
"extend",
"calibration",
"variants",
"takes",
"desired",
"relies",
"allows",
"equivalent",
"extremely",
"preliminary"
],
[
"gradient",
"functions",
"stochastic",
"optimization",
"function",
"convergence",
"local",
"convex",
"prior",
"non"
],
[
"differential",
"solving",
"constraint",
"equations",
"partial",
"stability",
"condition",
"gradients",
"inverse",
"equation"
],
[
"distribution",
"inference",
"estimation",
"bayesian",
"regression",
"distributions",
"gaussian",
"variables",
"sparse",
"processes"
],
[
"chosen",
"et",
"approximately",
"connections",
"al",
"mathematical",
"quantum",
"measurement",
"sense",
"characterization"
],
[
"fidelity",
"compression",
"parallel",
"capacity",
"expensive",
"speed",
"batch",
"costs",
"reducing",
"coding"
],
[
"attacks",
"adversarial",
"examples",
"privacy",
"detection",
"generative",
"attack",
"robustness",
"against",
"robust"
],
[
"class",
"sample",
"case",
"algorithm",
"random",
"linear",
"noise",
"matrix",
"sampling",
"selection"
],
[
"strongly",
"family",
"decomposition",
"matrices",
"recovery",
"kernels",
"recover",
"smooth",
"sensitivity",
"polynomial"
],
[
"vectors",
"fields",
"informative",
"arbitrary",
"estimated",
"core",
"propagation",
"constructed",
"limits",
"negative"
],
[
"iteration",
"posterior",
"variational",
"divergence",
"approximate",
"approximations",
"monte",
"markov",
"carlo",
"rewards"
],
[
"hyper",
"optimizing",
"tuning",
"compute",
"faster",
"magnitude",
"hyperparameters",
"residual",
"implement",
"sizes"
],
[
"propose",
"framework",
"novel",
"state",
"demonstrate",
"approach",
"learn",
"design",
"complex",
"real"
],
[
"natural",
"how",
"system",
"make",
"systems",
"sequence",
"long",
"behavior",
"like",
"about"
],
[
"targeting",
"reliability",
"incorporated",
"currently",
"affect",
"know",
"later",
"majority",
"detail",
"taken"
],
[
"selecting",
"running",
"criteria",
"contribution",
"established",
"unfortunately",
"depends",
"versus",
"produces",
"program"
],
[
"in",
"of",
"to",
"the",
"and",
"we",
"for",
"that",
"is",
"on"
],
[
"bias",
"content",
"overall",
"identifying",
"nature",
"collection",
"scores",
"ranking",
"protein",
"collected"
],
[
"true",
"selected",
"algorithmic",
"increases",
"produces",
"program",
"learners",
"ratio",
"select",
"contribution"
],
[
"takes",
"computationally",
"advantage",
"far",
"extend",
"motivated",
"allows",
"calibration",
"relies",
"handle"
],
[
"come",
"evaluating",
"third",
"behaviour",
"examine",
"turn",
"solely",
"offer",
"appropriate",
"adversary"
],
[
"clustering",
"latent",
"metric",
"embedding",
"representation",
"graphs",
"graph",
"structure",
"distance",
"nodes"
],
[
"estimating",
"probabilistic",
"conditional",
"tree",
"statistics",
"factorization",
"variation",
"flows",
"trees",
"graphical"
],
[
"analyzed",
"engineering",
"growing",
"review",
"science",
"tools",
"last",
"technologies",
"growth",
"risks"
],
[
"online",
"agent",
"strategy",
"rl",
"reinforcement",
"control",
"policy",
"decision",
"agents",
"strategies"
],
[
"high",
"computational",
"scale",
"parameters",
"efficient",
"low",
"large",
"up",
"memory",
"efficiency"
],
[
"variations",
"transferring",
"mapping",
"position",
"another",
"tend",
"auxiliary",
"mismatch",
"autoencoders",
"scenes"
],
[
"applicable",
"suggests",
"path",
"allows",
"experimentally",
"principled",
"minimal",
"feasibility",
"nn",
"importantly"
],
[
"surface",
"localization",
"mri",
"signal",
"consuming",
"magnetic",
"resonance",
"employed",
"sign",
"material"
],
[
"proposal",
"differentiable",
"autoregressive",
"details",
"employ",
"requiring",
"forward",
"direct",
"dropout",
"iteratively"
],
[
"filter",
"built",
"boosting",
"publicly",
"six",
"20",
"50",
"library",
"false",
"cells"
],
[
"image",
"convolutional",
"supervised",
"recognition",
"images",
"classifier",
"code",
"object",
"pre",
"visual"
],
[
"used",
"prediction",
"data",
"model",
"models",
"datasets",
"different",
"time",
"using",
"modeling"
],
[
"rules",
"shared",
"survey",
"reference",
"aspects",
"relational",
"recommender",
"outcomes",
"individuals",
"collaborative"
],
[
"machine",
"research",
"was",
"human",
"were",
"processing",
"developed",
"ml",
"quality",
"test"
]
] | 2,709.473035 | intfloat/e5-large-v2 | 0.97 | -0.247426 | 0.128677 | 0.876975 |
ArXiv ML Papers | S³_angular | 42 | 10 | [
[
"optimizer",
"focuses",
"specialized",
"compression",
"optimisation",
"preprocessing",
"optimize",
"motivation",
"visualization",
"adaptation"
],
[
"probability",
"uncertainty",
"diagnostic",
"bayesian",
"medical",
"disease",
"medicine",
"perturbations",
"diseases",
"interpretation"
],
[
"classifier",
"classifiers",
"benchmarks",
"optimisation",
"optimize",
"optimization",
"backpropagation",
"gpu",
"optimizes",
"optimizing"
],
[
"eye",
"label",
"taxonomy",
"attribute",
"classifying",
"classify",
"labels",
"classification",
"labeling",
"class"
],
[
"internet",
"github",
"add",
"streaming",
"computing",
"services",
"feed",
"gpus",
"scalability",
"module"
],
[
"nlp",
"sentences",
"logic",
"novel",
"words",
"sentence",
"hypothesis",
"dependencies",
"languages",
"linguistic"
],
[
"minimizing",
"approximating",
"minimize",
"simplified",
"approximation",
"approximations",
"limitations",
"simpler",
"weaker",
"theorem"
],
[
"nodes",
"graph",
"graphs",
"topology",
"network",
"vertices",
"node",
"networks",
"vertex",
"combinatorial"
],
[
"autoencoders",
"convolutional",
"convolutions",
"backpropagation",
"neural",
"imagenet",
"neuron",
"neurons",
"convolution",
"gnn"
],
[
"strategy",
"robot",
"reinforcement",
"learning",
"strategies",
"learners",
"robotics",
"robots",
"learnable",
"robotic"
]
] | 3.597447 | intfloat/e5-large-v2 | 0.97 | -0.333114 | 0.241362 | 0.873781 |
ArXiv ML Papers | S³_angular | 42 | 20 | [
[
"outperforming",
"empirically",
"kl",
"dl",
"outperformed",
"classifying",
"assessment",
"reliably",
"records",
"empirical"
],
[
"nodes",
"backbone",
"network",
"vertices",
"networks",
"vertex",
"graphs",
"node",
"graph",
"structure"
],
[
"behaviour",
"temporal",
"memory",
"patterns",
"behavior",
"spatiotemporal",
"eeg",
"time",
"ensembles",
"behavioral"
],
[
"gradients",
"rigorously",
"rigorous",
"smoothing",
"gradient",
"descent",
"consistently",
"optimization",
"penalty",
"benchmarks"
],
[
"dimensionality",
"dimensional",
"leverages",
"leveraged",
"reveals",
"tensor",
"explicitly",
"finite",
"embeddings",
"tensors"
],
[
"densities",
"density",
"statistical",
"dataset",
"compression",
"sizes",
"computes",
"weight",
"compressed",
"numerical"
],
[
"heuristics",
"tools",
"heuristic",
"tool",
"trees",
"analyze",
"analyzing",
"tree",
"evaluate",
"recursive"
],
[
"forecasting",
"forecasts",
"forecast",
"predicting",
"prediction",
"predictor",
"predicts",
"predict",
"predicted",
"predictive"
],
[
"sparse",
"poor",
"regret",
"fewer",
"bandits",
"bandit",
"unsupervised",
"greedy",
"limited",
"smaller"
],
[
"treatment",
"healthcare",
"medical",
"patients",
"diseases",
"medicine",
"health",
"diagnosis",
"healthy",
"disease"
],
[
"uncertainty",
"false",
"measurement",
"measurements",
"capacity",
"estimation",
"statistics",
"uncertainties",
"never",
"healthcare"
],
[
"english",
"languages",
"speech",
"textual",
"translation",
"voice",
"language",
"sentence",
"sentences",
"textit"
],
[
"observation",
"remain",
"observations",
"regularization",
"keeping",
"dirichlet",
"generalization",
"retaining",
"observational",
"discriminative"
],
[
"overhead",
"vehicles",
"vehicle",
"grounded",
"distance",
"transportation",
"transport",
"sensor",
"tracking",
"technical"
],
[
"concepts",
"mixed",
"lot",
"compact",
"converges",
"capacity",
"compression",
"converge",
"hypothesis",
"allocation"
],
[
"autoencoders",
"autoencoder",
"lstm",
"storage",
"dnn",
"pytorch",
"backpropagation",
"dnns",
"storing",
"bayesian"
],
[
"subspace",
"rank",
"factorization",
"matrices",
"subspaces",
"tensor",
"tensors",
"matrix",
"compress",
"linear"
],
[
"ensembles",
"overfitting",
"outperform",
"pretrained",
"outperforms",
"outperforming",
"outperformed",
"generalizability",
"unlabeled",
"subset"
],
[
"backpropagation",
"agnostic",
"optimum",
"gnn",
"gnns",
"denoising",
"nonlinear",
"input",
"calibration",
"inputs"
],
[
"neurons",
"imagenet",
"convolutions",
"convolutional",
"backpropagation",
"deep",
"deeper",
"cnn",
"dnn",
"convolution"
]
] | 3.776158 | intfloat/e5-large-v2 | 0.945 | -0.319573 | 0.218771 | 0.901801 |
ArXiv ML Papers | S³_angular | 42 | 30 | [
[
"dnn",
"margin",
"total",
"deep",
"derivatives",
"proof",
"deeper",
"mathcal",
"dnns",
"theorem"
],
[
"ensemble",
"synthesis",
"extracts",
"distillation",
"factorization",
"batch",
"experimental",
"effects",
"factor",
"experimentally"
],
[
"forecast",
"predictor",
"forecasts",
"predicting",
"forecasting",
"predictors",
"predicts",
"prediction",
"weather",
"predictions"
],
[
"computes",
"weight",
"dataset",
"textual",
"approximated",
"datasets",
"outlier",
"approximates",
"textit",
"databases"
],
[
"gan",
"generate",
"image",
"gans",
"generates",
"generating",
"generative",
"generated",
"visual",
"art"
],
[
"besides",
"considers",
"relies",
"suggests",
"consider",
"applies",
"suggesting",
"although",
"utilizes",
"useful"
],
[
"autoregressive",
"causal",
"time",
"past",
"temporal",
"latency",
"explainable",
"spatiotemporal",
"recurrent",
"records"
],
[
"formalism",
"manifold",
"shift",
"fusion",
"shifts",
"physics",
"coupling",
"functional",
"alignment",
"invariance"
],
[
"acoustic",
"performances",
"speaker",
"speech",
"audio",
"sound",
"voice",
"recorded",
"frequency",
"performing"
],
[
"together",
"classes",
"diversity",
"taxonomy",
"classification",
"class",
"stacked",
"multi",
"boxes",
"each"
],
[
"measured",
"compatible",
"verified",
"practical",
"verify",
"reliable",
"measurements",
"implementations",
"survey",
"requirements"
],
[
"adversarial",
"attacker",
"malware",
"attack",
"adversaries",
"attacks",
"adversarially",
"malicious",
"adversary",
"vulnerability"
],
[
"inequality",
"fairness",
"equally",
"biases",
"biased",
"bias",
"preferences",
"preference",
"discrimination",
"discriminate"
],
[
"learners",
"learnable",
"trainable",
"predict",
"predictor",
"backpropagation",
"lstm",
"learner",
"predicts",
"classifier"
],
[
"health",
"healthcare",
"medicine",
"medical",
"treatment",
"clinical",
"patients",
"diseases",
"disease",
"patient"
],
[
"popularity",
"suggest",
"personalized",
"like",
"recommender",
"recommendations",
"recommendation",
"popular",
"user",
"employ"
],
[
"transferring",
"adapts",
"adapt",
"adapting",
"transfer",
"pretrained",
"transferred",
"retrieval",
"adapted",
"overfitting"
],
[
"bayesian",
"bayes",
"probability",
"hyperparameters",
"covariates",
"probabilistic",
"gaussian",
"likelihood",
"probabilities",
"hypothesis"
],
[
"estimating",
"nn",
"parameter",
"parameters",
"calibration",
"backpropagation",
"uncertainty",
"estimation",
"estimator",
"neural"
],
[
"label",
"supervised",
"unlabeled",
"errors",
"supervision",
"error",
"uncertain",
"reliable",
"denoising",
"labels"
],
[
"forest",
"dataset",
"datasets",
"github",
"map",
"patch",
"python",
"poor",
"risks",
"patches"
],
[
"observation",
"observations",
"differential",
"learn",
"observational",
"learned",
"learner",
"nonparametric",
"learners",
"differentiable"
],
[
"algorithms",
"gradient",
"accelerate",
"acceleration",
"faster",
"running",
"consistently",
"gradients",
"goal",
"runs"
],
[
"dnn",
"dnns",
"nonnegative",
"pairwise",
"improves",
"orthogonal",
"gnns",
"cnns",
"backpropagation",
"discriminative"
],
[
"collaborative",
"distributed",
"decentralized",
"cooperative",
"centralized",
"federated",
"mutual",
"sharing",
"shared",
"jointly"
],
[
"autoencoder",
"autoencoders",
"coding",
"encode",
"encodes",
"encoding",
"encoded",
"encoder",
"encoders",
"compression"
],
[
"substantially",
"previously",
"very",
"regarding",
"much",
"yet",
"drastically",
"unfortunately",
"poorly",
"rarely"
],
[
"reaching",
"convergence",
"evolving",
"evolution",
"reach",
"distributions",
"infinite",
"divergence",
"streams",
"flows"
],
[
"translation",
"sentences",
"semantic",
"sentence",
"textual",
"linguistic",
"semantically",
"language",
"translate",
"english"
],
[
"ai",
"kernels",
"corpus",
"processing",
"preprocessing",
"fields",
"technology",
"field",
"softmax",
"sqrt"
]
] | 4.023511 | intfloat/e5-large-v2 | 0.96 | -0.29823 | 0.230479 | 0.896805 |
ArXiv ML Papers | S³_angular | 42 | 40 | [
[
"found",
"detection",
"detect",
"find",
"inspection",
"look",
"detecting",
"investigation",
"source",
"enforce"
],
[
"entity",
"provide",
"factor",
"energy",
"component",
"evaluate",
"components",
"element",
"presenting",
"hand"
],
[
"nodes",
"graph",
"networks",
"network",
"graphs",
"vertices",
"vertex",
"node",
"graphical",
"edges"
],
[
"synthetic",
"observation",
"separately",
"demonstrating",
"encoders",
"simultaneous",
"decoding",
"observational",
"developments",
"mcmc"
],
[
"adversarially",
"malware",
"attacks",
"attacker",
"attack",
"adversaries",
"adversarial",
"threat",
"malicious",
"adversary"
],
[
"attention",
"focuses",
"focusing",
"focus",
"focused",
"actively",
"interest",
"aware",
"vision",
"add"
],
[
"strategies",
"market",
"bandits",
"costly",
"bandit",
"exploiting",
"online",
"regret",
"greedy",
"offers"
],
[
"mismatch",
"diverse",
"contrary",
"noise",
"fashion",
"noisy",
"vehicles",
"videos",
"static",
"forces"
],
[
"neural",
"perception",
"backpropagation",
"imagenet",
"exploration",
"clustering",
"svm",
"classifier",
"classifiers",
"hessian"
],
[
"deep",
"dnn",
"dnns",
"deeper",
"backpropagation",
"highest",
"difficulties",
"bernoulli",
"double",
"problems"
],
[
"privacy",
"embedding",
"private",
"hidden",
"publicly",
"embeddings",
"security",
"synthetic",
"reveal",
"restricted"
],
[
"benchmarks",
"benchmark",
"analytics",
"kl",
"dimensionality",
"metrics",
"hyperparameters",
"capacity",
"outperformed",
"classification"
],
[
"ignore",
"tensor",
"split",
"pytorch",
"workflow",
"batch",
"stochastic",
"component",
"tensors",
"share"
],
[
"indicating",
"inferring",
"predicting",
"guidance",
"residual",
"imputation",
"predictive",
"predict",
"avoiding",
"programming"
],
[
"tight",
"stochastic",
"bottleneck",
"gradients",
"sgd",
"edge",
"convex",
"convexity",
"gradient",
"fast"
],
[
"count",
"pixel",
"scalability",
"correlated",
"markov",
"pixels",
"counterpart",
"large",
"showed",
"vast"
],
[
"fairness",
"discriminate",
"discrimination",
"bias",
"inequality",
"biased",
"equally",
"fair",
"equal",
"biases"
],
[
"source",
"observations",
"teacher",
"learning",
"knowledge",
"phenomenon",
"sources",
"entropy",
"learnt",
"separated"
],
[
"range",
"simulations",
"exposure",
"bias",
"ranging",
"experiment",
"lies",
"sampling",
"accuracy",
"experiments"
],
[
"uncertain",
"bayes",
"gaussian",
"probability",
"probabilities",
"probabilistic",
"priors",
"likelihood",
"particle",
"bayesian"
],
[
"coding",
"encode",
"encoded",
"encoder",
"encoders",
"autoencoder",
"autoencoders",
"encodes",
"encoding",
"vae"
],
[
"sentence",
"translation",
"english",
"languages",
"sentences",
"words",
"linguistic",
"speech",
"language",
"texts"
],
[
"interpretation",
"short",
"contrastive",
"critic",
"feature",
"characterization",
"longer",
"careful",
"indicate",
"word"
],
[
"approximation",
"partition",
"interpolation",
"convolution",
"convolutions",
"partitioning",
"derivatives",
"convolutional",
"approximations",
"mixture"
],
[
"transfer",
"transferred",
"transmission",
"adapts",
"transferring",
"adapt",
"shift",
"sharing",
"attributed",
"adapting"
],
[
"audio",
"speech",
"acoustic",
"voice",
"sound",
"speaker",
"recorded",
"communications",
"performing",
"frequencies"
],
[
"robust",
"objects",
"equipped",
"object",
"database",
"robustness",
"sensors",
"grounded",
"surface",
"coupled"
],
[
"major",
"selection",
"inspection",
"concern",
"chosen",
"unlabeled",
"selecting",
"minimum",
"mostly",
"partial"
],
[
"rank",
"subspaces",
"subspace",
"factorization",
"linear",
"matrices",
"sparse",
"matrix",
"resolution",
"rl"
],
[
"devise",
"implement",
"software",
"python",
"github",
"patches",
"patch",
"forest",
"vanilla",
"development"
],
[
"movement",
"moving",
"motion",
"brain",
"step",
"neural",
"tracking",
"head",
"strictly",
"steps"
],
[
"yet",
"consequently",
"furthermore",
"locally",
"regarding",
"unfortunately",
"previously",
"actual",
"low",
"subsequently"
],
[
"backpropagation",
"rnn",
"gnns",
"pretrained",
"rnns",
"lstm",
"boosting",
"xgboost",
"dnns",
"enhances"
],
[
"besides",
"additionally",
"considers",
"suggests",
"utilizes",
"although",
"especially",
"introduces",
"applies",
"some"
],
[
"dynamical",
"manifold",
"differentially",
"stable",
"automatically",
"laplacian",
"finite",
"automated",
"turns",
"spectral"
],
[
"hessian",
"similarities",
"interplay",
"similarity",
"parallel",
"hyper",
"potential",
"utilization",
"versus",
"virtual"
],
[
"track",
"algorithms",
"streams",
"runtime",
"streaming",
"stream",
"compute",
"computes",
"algorithmic",
"program"
],
[
"controller",
"reinforcement",
"policy",
"rl",
"conditioning",
"control",
"policies",
"experience",
"play",
"simulation"
],
[
"interpretable",
"latent",
"explainable",
"observable",
"interpretability",
"operators",
"task",
"dnn",
"operator",
"imagenet"
],
[
"cluster",
"clustering",
"segment",
"partitioning",
"segmentation",
"clusters",
"decentralized",
"localized",
"region",
"occurs"
]
] | 4.075726 | intfloat/e5-large-v2 | 0.955 | -0.325799 | 0.192967 | 0.894473 |
ArXiv ML Papers | S³_angular | 42 | 50 | [
[
"brain",
"relative",
"form",
"shape",
"tail",
"motion",
"posterior",
"neuron",
"neurons",
"thus"
],
[
"vehicle",
"vehicles",
"car",
"transport",
"navigation",
"traffic",
"driving",
"gps",
"transportation",
"auto"
],
[
"static",
"mismatch",
"xgboost",
"compositional",
"contrary",
"covariance",
"forces",
"coupling",
"invariance",
"combines"
],
[
"subspaces",
"energy",
"entropy",
"quantum",
"physics",
"quantized",
"tensors",
"tensor",
"quantitative",
"quantization"
],
[
"expressive",
"augmentation",
"iterations",
"iteration",
"robustness",
"invariant",
"adaptively",
"trajectory",
"smoothness",
"recurrent"
],
[
"matching",
"estimators",
"authors",
"mismatch",
"fitting",
"computes",
"mathbb",
"predicts",
"single",
"datasets"
],
[
"complicated",
"entities",
"category",
"intractable",
"batch",
"complexities",
"degradation",
"hierarchy",
"component",
"complexity"
],
[
"besides",
"considers",
"suggests",
"encourages",
"establish",
"approaches",
"utilizes",
"introduces",
"although",
"additionally"
],
[
"cluster",
"clustering",
"partitioning",
"clusters",
"segmentation",
"groups",
"segment",
"supervision",
"association",
"partition"
],
[
"optimum",
"pixel",
"priori",
"accessible",
"minimal",
"minimum",
"insufficient",
"evaluated",
"softmax",
"lightweight"
],
[
"autoencoders",
"autoencoder",
"svm",
"encoder",
"vae",
"encoders",
"lstm",
"autoregressive",
"variational",
"devise"
],
[
"adversaries",
"attacks",
"attack",
"adversarial",
"adversarially",
"attacker",
"threat",
"adversary",
"malicious",
"vulnerable"
],
[
"models",
"model",
"transformers",
"dataset",
"datasets",
"transformer",
"objects",
"modelling",
"manifold",
"modeling"
],
[
"differential",
"differentially",
"dynamical",
"observational",
"dynamically",
"analytical",
"claim",
"analytically",
"laplacian",
"imply"
],
[
"identity",
"classified",
"privacy",
"private",
"hidden",
"publicly",
"extracted",
"reveal",
"embedding",
"pass"
],
[
"expressed",
"semantic",
"manifold",
"grounded",
"intelligent",
"intrinsic",
"nonconvex",
"flexible",
"dimensionality",
"topological"
],
[
"dedicated",
"designs",
"loop",
"subspace",
"challenging",
"embedded",
"cycles",
"turn",
"exploited",
"incorporated"
],
[
"ask",
"remote",
"dependencies",
"huge",
"expand",
"exploring",
"used",
"walk",
"numerous",
"former"
],
[
"kl",
"benchmarks",
"contain",
"sensitivity",
"handling",
"containing",
"bottleneck",
"hyperparameters",
"hyperparameter",
"operators"
],
[
"mixture",
"encoding",
"stochastic",
"diffusion",
"aggregate",
"mixed",
"encoded",
"information",
"overview",
"flow"
],
[
"eeg",
"wearable",
"behaviour",
"static",
"stationary",
"sensor",
"sensors",
"activity",
"behavioral",
"devices"
],
[
"proposal",
"compatible",
"solutions",
"guarantee",
"solving",
"deal",
"guaranteed",
"solvers",
"solver",
"reliable"
],
[
"degrade",
"imaging",
"mri",
"degradation",
"imagenet",
"overfitting",
"filters",
"preprocessing",
"pixels",
"localization"
],
[
"distillation",
"learned",
"learnt",
"knowledge",
"learning",
"learner",
"learn",
"intelligence",
"learners",
"studying"
],
[
"measurements",
"viable",
"measured",
"practical",
"measurement",
"methodology",
"variability",
"step",
"measure",
"metrics"
],
[
"hyperparameter",
"partition",
"hierarchical",
"hyper",
"partitioning",
"scalable",
"intensive",
"parametric",
"potential",
"consideration"
],
[
"remove",
"black",
"carefully",
"previous",
"year",
"scratch",
"removing",
"before",
"revisit",
"considers"
],
[
"ml",
"computer",
"management",
"machine",
"machines",
"mining",
"industrial",
"ai",
"managing",
"industry"
],
[
"dynamics",
"discrete",
"screening",
"denoising",
"controller",
"separation",
"smoothness",
"distillation",
"convolutions",
"convolution"
],
[
"classifiers",
"labeling",
"svm",
"automated",
"classifier",
"automatically",
"algorithmic",
"embedded",
"textit",
"label"
],
[
"perception",
"sampling",
"sensing",
"representative",
"detection",
"detected",
"susceptible",
"detect",
"detecting",
"selection"
],
[
"weather",
"forest",
"forecast",
"evolution",
"forecasting",
"forecasts",
"evolving",
"ensembles",
"cloud",
"outside"
],
[
"yield",
"compressed",
"preserving",
"vast",
"heavy",
"sparse",
"long",
"extraction",
"large",
"far"
],
[
"recommendation",
"recommender",
"recommendations",
"suggested",
"suggest",
"personalized",
"popular",
"popularity",
"preferences",
"influence"
],
[
"points",
"heuristic",
"inference",
"infinite",
"inherent",
"deployment",
"computations",
"reasoning",
"analytics",
"heuristics"
],
[
"impractical",
"ratio",
"sensor",
"sensors",
"unlike",
"exposure",
"low",
"exchange",
"differences",
"less"
],
[
"epsilon",
"limits",
"boundary",
"limit",
"converges",
"boundaries",
"approximations",
"offers",
"applied",
"capacity"
],
[
"generalization",
"generalizing",
"generalizability",
"generalizes",
"generalize",
"fewer",
"characterizing",
"asymptotically",
"typical",
"rarely"
],
[
"principle",
"minimax",
"taxonomy",
"satisfies",
"corpus",
"fields",
"posterior",
"softmax",
"straightforward",
"sqrt"
],
[
"strict",
"coarse",
"rigorous",
"presence",
"cell",
"granularity",
"concerns",
"exists",
"rigorously",
"sublinear"
],
[
"platforms",
"boxes",
"dimensions",
"box",
"description",
"modes",
"field",
"concrete",
"diverse",
"mode"
],
[
"storage",
"compression",
"streaming",
"storing",
"records",
"entropy",
"algorithmic",
"vital",
"relevance",
"sequence"
],
[
"linear",
"minimax",
"ml",
"multimodal",
"nonlinear",
"mcmc",
"linearly",
"mismatch",
"resnet",
"93"
],
[
"rapid",
"randomly",
"diffusion",
"randomized",
"manifold",
"gaussian",
"svm",
"mnist",
"stochastic",
"random"
],
[
"biological",
"protein",
"mechanism",
"molecular",
"genetic",
"biology",
"interactions",
"beneficial",
"mechanisms",
"chemical"
],
[
"modeling",
"accessible",
"models",
"representations",
"representation",
"modelling",
"modeled",
"model",
"represented",
"estimators"
],
[
"applicability",
"adopted",
"configuration",
"count",
"variants",
"verified",
"variant",
"compatible",
"markov",
"categorical"
],
[
"explaining",
"interpretability",
"aspects",
"aspect",
"interpretable",
"explanations",
"interpretation",
"meaningful",
"explainable",
"interpret"
],
[
"agnostic",
"dnn",
"tuning",
"dnns",
"nn",
"gnns",
"gnn",
"dataset",
"classifier",
"outlier"
],
[
"bandits",
"regret",
"bandit",
"greedy",
"naive",
"strategies",
"opportunities",
"exploiting",
"strategy",
"ucb"
]
] | 4.625049 | intfloat/e5-large-v2 | 0.93 | -0.364717 | 0.183481 | 0.915853 |
ArXiv ML Papers | S³_combined | 42 | 10 | [
[
"spatial",
"mapping",
"optimisation",
"preprocessing",
"compression",
"focuses",
"optimizer",
"optimizes",
"oriented",
"visualization"
],
[
"medical",
"diagnostic",
"bayesian",
"perturbations",
"disease",
"uncertainty",
"diseases",
"poisoning",
"probability",
"perturbation"
],
[
"optimizing",
"optimization",
"benchmarks",
"gpu",
"classifiers",
"classifier",
"optimizes",
"optimize",
"optimisation",
"outperforming"
],
[
"eye",
"class",
"taxonomy",
"face",
"cross",
"label",
"classification",
"classify",
"classifying",
"labels"
],
[
"federated",
"network",
"networks",
"streaming",
"node",
"feed",
"internet",
"deployed",
"iot",
"social"
],
[
"novel",
"dependencies",
"linguistic",
"hypothesis",
"nlp",
"logic",
"words",
"sentence",
"sentences",
"languages"
],
[
"solves",
"solvers",
"optimizes",
"besides",
"minimize",
"differentiable",
"simplified",
"constrained",
"especially",
"minimizing"
],
[
"graph",
"graphs",
"topology",
"nodes",
"network",
"vertices",
"topological",
"networks",
"vertex",
"node"
],
[
"gnn",
"neurons",
"neural",
"convolutional",
"convolutions",
"neuron",
"imagenet",
"backpropagation",
"dnn",
"abstraction"
],
[
"robotic",
"strategy",
"learning",
"autonomous",
"reinforcement",
"robotics",
"strategies",
"robots",
"experience",
"exploration"
]
] | 3.557533 | intfloat/e5-large-v2 | 0.94 | -0.296069 | 0.237198 | 0.861545 |
ArXiv ML Papers | S³_combined | 42 | 20 | [
[
"outperforming",
"reliably",
"nearest",
"empirical",
"empirically",
"classifying",
"assessment",
"dl",
"kl",
"distances"
],
[
"graph",
"backbone",
"nodes",
"network",
"networks",
"vertex",
"node",
"vertices",
"graphs",
"graphical"
],
[
"behaviour",
"recurrent",
"time",
"eeg",
"behavior",
"spatiotemporal",
"behaviors",
"temporal",
"behavioral",
"transient"
],
[
"rigorous",
"gradient",
"smoothing",
"rigorously",
"gradients",
"penalty",
"smooth",
"regularized",
"consistently",
"optimization"
],
[
"leverages",
"dimensional",
"dimensionality",
"leveraged",
"relations",
"uncertainties",
"explicitly",
"convexity",
"tensor",
"leverage"
],
[
"densities",
"statistical",
"cluster",
"density",
"compressed",
"computes",
"sizes",
"clusters",
"remains",
"texts"
],
[
"tree",
"tools",
"sketch",
"heuristic",
"sequence",
"sequences",
"heuristics",
"investigates",
"trees",
"evaluate"
],
[
"forecasting",
"forecasts",
"predicts",
"predicting",
"forecast",
"predictor",
"predict",
"autoregressive",
"predicted",
"predictors"
],
[
"sparse",
"bandits",
"bandit",
"regret",
"greedy",
"smaller",
"fewer",
"poor",
"small",
"unsupervised"
],
[
"diseases",
"disease",
"patients",
"medical",
"healthcare",
"health",
"medicine",
"patient",
"treatment",
"diagnosis"
],
[
"never",
"uncertainties",
"uncertainty",
"measurements",
"false",
"capacity",
"measurement",
"estimation",
"insufficient",
"statistics"
],
[
"audio",
"english",
"translation",
"voice",
"textual",
"speech",
"language",
"textit",
"languages",
"translate"
],
[
"keeping",
"regularization",
"observations",
"observational",
"observation",
"remain",
"retaining",
"outliers",
"regularized",
"phenomenon"
],
[
"transport",
"grounded",
"distance",
"overhead",
"vehicles",
"camera",
"sensor",
"vehicle",
"transportation",
"gps"
],
[
"spatial",
"coordinate",
"converges",
"converge",
"mixed",
"compact",
"capacity",
"continuous",
"convergence",
"hypothesis"
],
[
"dnns",
"lstm",
"autoencoders",
"dnn",
"autoencoder",
"backpropagation",
"rnns",
"bayesian",
"storage",
"storing"
],
[
"matrices",
"matrix",
"tensor",
"rank",
"subspaces",
"factorization",
"subspace",
"linear",
"tensors",
"sparse"
],
[
"pretrained",
"outperform",
"outperformed",
"outperforms",
"overfitting",
"unlabeled",
"ensembles",
"boosting",
"outperforming",
"subset"
],
[
"backpropagation",
"optimum",
"gnns",
"gnn",
"uncertainties",
"nonlinear",
"inputs",
"tuned",
"rnns",
"calibration"
],
[
"deeper",
"convolutions",
"convolutional",
"neuron",
"neurons",
"cnn",
"backpropagation",
"dnn",
"imagenet",
"deep"
]
] | 3.821646 | intfloat/e5-large-v2 | 0.945 | -0.318014 | 0.217034 | 0.900606 |
ArXiv ML Papers | S³_combined | 42 | 30 | [
[
"dnn",
"mathcal",
"margin",
"derivatives",
"dnns",
"total",
"deeper",
"deep",
"theorem",
"proof"
],
[
"generates",
"factor",
"distillation",
"extracts",
"synthesis",
"effects",
"perturbation",
"ensembles",
"ensemble",
"factorization"
],
[
"forecasts",
"forecast",
"forecasting",
"predicts",
"predictor",
"weather",
"predicting",
"predictions",
"predictors",
"predicted"
],
[
"weight",
"texts",
"outlier",
"textual",
"remains",
"approximates",
"textit",
"computes",
"approximated",
"estimates"
],
[
"gan",
"generative",
"generates",
"generate",
"gans",
"generating",
"graphical",
"generated",
"art",
"image"
],
[
"consider",
"interesting",
"several",
"involves",
"although",
"considers",
"considering",
"besides",
"suggests",
"discusses"
],
[
"temporal",
"spatiotemporal",
"autoregressive",
"causal",
"recurrent",
"explainable",
"time",
"transient",
"past",
"times"
],
[
"fusion",
"formalism",
"manifold",
"alignment",
"shift",
"flows",
"invariance",
"flow",
"physics",
"invariant"
],
[
"audio",
"recorded",
"speech",
"voice",
"speaker",
"acoustic",
"sound",
"frequency",
"frequencies",
"communications"
],
[
"together",
"classification",
"multi",
"stacked",
"taxonomy",
"diversity",
"classes",
"widespread",
"classify",
"class"
],
[
"measured",
"verify",
"verified",
"compatible",
"practical",
"reliable",
"measurements",
"survey",
"challenge",
"propose"
],
[
"adversarially",
"attacker",
"malicious",
"adversarial",
"attack",
"malware",
"attacks",
"adversaries",
"adversary",
"security"
],
[
"bias",
"discrimination",
"biases",
"biased",
"fairness",
"discriminate",
"preferences",
"preference",
"equally",
"inequality"
],
[
"learners",
"predicts",
"predict",
"predictor",
"lstm",
"learnable",
"trainable",
"backpropagation",
"learner",
"classifier"
],
[
"medical",
"healthcare",
"medicine",
"disease",
"diseases",
"health",
"patients",
"treatment",
"patient",
"clinical"
],
[
"recommendation",
"recommendations",
"personalized",
"recommender",
"popularity",
"like",
"suggest",
"popular",
"employ",
"seek"
],
[
"transferring",
"adapt",
"adapts",
"transferred",
"transfer",
"retrieval",
"adapting",
"adapted",
"overfitting",
"augmentation"
],
[
"bayes",
"gaussian",
"probabilistic",
"probabilities",
"likelihood",
"probability",
"hyperparameters",
"variability",
"covariates",
"bayesian"
],
[
"uncertainty",
"parameter",
"neural",
"uncertainties",
"nn",
"parameters",
"estimation",
"constrain",
"calibration",
"constraint"
],
[
"uncertain",
"detector",
"label",
"supervision",
"supervised",
"errors",
"denoising",
"unlabeled",
"noisy",
"reliable"
],
[
"poor",
"map",
"risks",
"urban",
"sketch",
"patch",
"forest",
"patches",
"github",
"landscape"
],
[
"learner",
"differentiable",
"regularization",
"learners",
"differential",
"learns",
"observation",
"observations",
"differentially",
"learned"
],
[
"consistently",
"gradients",
"acceleration",
"smoothing",
"run",
"running",
"runs",
"gradient",
"goal",
"convex"
],
[
"dnn",
"dnns",
"orthogonal",
"gnns",
"nonnegative",
"cnns",
"gnn",
"rnn",
"enhances",
"discriminator"
],
[
"federated",
"collaborative",
"centralized",
"distributed",
"cooperative",
"decentralized",
"sharing",
"mutual",
"shared",
"together"
],
[
"coding",
"autoencoders",
"autoencoder",
"encode",
"encoded",
"encoder",
"encodes",
"encoding",
"encoders",
"compression"
],
[
"drastically",
"very",
"uncertainties",
"unfortunately",
"poorly",
"much",
"regarding",
"yet",
"exposure",
"leverages"
],
[
"flows",
"reach",
"evolving",
"convergence",
"evolution",
"distributions",
"reaching",
"imitation",
"divergence",
"nn"
],
[
"translation",
"translate",
"sentences",
"linguistic",
"sentence",
"semantically",
"semantic",
"textual",
"language",
"languages"
],
[
"corpus",
"ai",
"kernels",
"fields",
"posterior",
"field",
"preprocessing",
"processing",
"sqrt",
"softmax"
]
] | 4.147288 | intfloat/e5-large-v2 | 0.956667 | -0.318527 | 0.218008 | 0.90568 |
ArXiv ML Papers | S³_combined | 42 | 40 | [
[
"detect",
"look",
"detection",
"investigation",
"found",
"detecting",
"inspection",
"find",
"enforce",
"investigates"
],
[
"ensemble",
"provide",
"component",
"components",
"energy",
"factor",
"hand",
"evaluate",
"element",
"function"
],
[
"vertex",
"graph",
"graphs",
"node",
"nodes",
"vertices",
"network",
"graphical",
"networks",
"edges"
],
[
"encoder",
"encoders",
"mcmc",
"simultaneous",
"observation",
"demonstrating",
"reconstruction",
"synthetic",
"separately",
"decoding"
],
[
"attacker",
"adversaries",
"adversarial",
"robustness",
"adversarially",
"attack",
"attacks",
"malware",
"threat",
"malicious"
],
[
"actively",
"aware",
"focusing",
"attention",
"focus",
"focused",
"focuses",
"interest",
"add",
"forgetting"
],
[
"regret",
"offers",
"bandits",
"bandit",
"greedy",
"costly",
"exploiting",
"exploration",
"online",
"market"
],
[
"static",
"mismatch",
"contrary",
"fashion",
"noisy",
"noise",
"vehicles",
"diverse",
"videos",
"forces"
],
[
"svm",
"exploration",
"neural",
"gradients",
"imagenet",
"clusters",
"clustering",
"cluster",
"classifier",
"perception"
],
[
"dnn",
"dnns",
"difficulty",
"deep",
"backpropagation",
"deeper",
"problems",
"difficulties",
"bernoulli",
"neural"
],
[
"embeddings",
"privacy",
"embedding",
"private",
"hidden",
"security",
"synthetic",
"publicly",
"restricted",
"reveal"
],
[
"benchmarks",
"dimensionality",
"kl",
"benchmark",
"classification",
"metrics",
"analytics",
"manifold",
"capacity",
"hyperparameters"
],
[
"batch",
"stochastic",
"split",
"ignore",
"simultaneous",
"share",
"tensors",
"workflow",
"tensor",
"pytorch"
],
[
"imputation",
"indicating",
"inferring",
"predicting",
"guidance",
"residual",
"predictive",
"delay",
"predict",
"predicted"
],
[
"convex",
"gradient",
"convexity",
"stochastic",
"nonlinear",
"tight",
"sgd",
"bottleneck",
"gradients",
"edge"
],
[
"counterpart",
"scalability",
"pixels",
"vast",
"showed",
"pixel",
"correlated",
"count",
"markov",
"large"
],
[
"discriminate",
"biased",
"bias",
"discrimination",
"fairness",
"biases",
"fair",
"inequality",
"equally",
"equal"
],
[
"forgetting",
"observations",
"knowledge",
"phenomenon",
"entropy",
"source",
"sources",
"learning",
"absence",
"decompose"
],
[
"ranging",
"bias",
"simulations",
"experiment",
"exposure",
"range",
"sampling",
"biases",
"experiments",
"accuracy"
],
[
"likelihood",
"bayes",
"probabilistic",
"probabilities",
"probability",
"gaussian",
"bayesian",
"uncertain",
"priors",
"posterior"
],
[
"autoencoders",
"autoencoder",
"encoder",
"encoded",
"encode",
"generative",
"coding",
"vae",
"encodes",
"encoders"
],
[
"sentences",
"sentence",
"linguistic",
"words",
"english",
"translation",
"language",
"languages",
"texts",
"textual"
],
[
"short",
"critic",
"characterization",
"decreases",
"contrastive",
"longer",
"attractive",
"careful",
"indicate",
"decrease"
],
[
"partition",
"partitioning",
"interpolation",
"derivatives",
"convolution",
"approximation",
"mixture",
"filtering",
"approximations",
"convolutions"
],
[
"transfer",
"adapt",
"transferring",
"transferred",
"adapts",
"transmission",
"sharing",
"shift",
"adapting",
"attributed"
],
[
"sound",
"voice",
"acoustic",
"recorded",
"audio",
"speech",
"speaker",
"noise",
"communications",
"frequencies"
],
[
"invariant",
"uncertainties",
"sensors",
"robustness",
"robust",
"grounded",
"poses",
"objects",
"object",
"equipped"
],
[
"inspection",
"unlabeled",
"minimal",
"focus",
"selection",
"minimum",
"selecting",
"chosen",
"pixel",
"designing"
],
[
"sparse",
"subspaces",
"matrices",
"subspace",
"resolution",
"matrix",
"factorization",
"rank",
"linear",
"tensors"
],
[
"devise",
"github",
"patches",
"forest",
"patch",
"implement",
"python",
"tree",
"aid",
"framework"
],
[
"motion",
"brain",
"moving",
"movement",
"step",
"neural",
"head",
"aiming",
"strictly",
"neurons"
],
[
"yet",
"regarding",
"consequently",
"subsequent",
"responses",
"unfortunately",
"prone",
"locally",
"neighbors",
"furthermore"
],
[
"dnns",
"boosting",
"backpropagation",
"rnn",
"rnns",
"gnns",
"xgboost",
"enhances",
"gnn",
"lstm"
],
[
"learns",
"besides",
"interesting",
"although",
"considers",
"additionally",
"attracted",
"increases",
"utilizes",
"especially"
],
[
"laplacian",
"dynamical",
"manifold",
"differentially",
"spectral",
"finite",
"automatically",
"automated",
"stable",
"turns"
],
[
"derive",
"pooling",
"similarities",
"versus",
"hyper",
"parallel",
"exploitation",
"interplay",
"intensive",
"hessian"
],
[
"streaming",
"stream",
"streams",
"algorithms",
"algorithmic",
"compute",
"track",
"runtime",
"computes",
"collection"
],
[
"play",
"rl",
"reinforcement",
"policy",
"conditioning",
"control",
"controller",
"environments",
"policies",
"agents"
],
[
"latent",
"dnn",
"explainable",
"task",
"interpretability",
"interpretable",
"observable",
"dnns",
"artifacts",
"subjects"
],
[
"clusters",
"clustering",
"cluster",
"partitioning",
"segmentation",
"segment",
"decentralized",
"communities",
"localized",
"finds"
]
] | 4.011805 | intfloat/e5-large-v2 | 0.9325 | -0.31289 | 0.192761 | 0.887805 |
ArXiv ML Papers | S³_combined | 42 | 50 | [
[
"relative",
"posterior",
"motion",
"tail",
"shape",
"brain",
"thus",
"neurons",
"neuron",
"form"
],
[
"traffic",
"vehicles",
"transportation",
"transport",
"driving",
"vehicle",
"car",
"gps",
"navigation",
"trajectory"
],
[
"mismatch",
"consistent",
"static",
"invariance",
"covariance",
"coupling",
"contrary",
"forces",
"compositional",
"xgboost"
],
[
"tensor",
"quantization",
"quantized",
"quantum",
"tensors",
"physics",
"quantitative",
"subspaces",
"energy",
"entropy"
],
[
"recurrent",
"robust",
"expressive",
"invariant",
"trajectory",
"robustness",
"augmentation",
"iteration",
"iterations",
"uncertain"
],
[
"mismatch",
"matching",
"mathbb",
"predicts",
"computes",
"authors",
"models",
"fitting",
"estimators",
"single"
],
[
"degradation",
"complexity",
"complicated",
"complexities",
"batch",
"category",
"co",
"hierarchy",
"decomposition",
"tensor"
],
[
"establish",
"introduces",
"encourages",
"solves",
"suggests",
"helps",
"some",
"established",
"attracted",
"besides"
],
[
"partitioning",
"supervision",
"clusters",
"clustering",
"cluster",
"segment",
"segmentation",
"groups",
"association",
"community"
],
[
"optimum",
"sparse",
"minimal",
"accessible",
"minimum",
"pixel",
"maximize",
"focus",
"lightweight",
"optimized"
],
[
"autoencoders",
"autoencoder",
"encoder",
"vae",
"encoders",
"variational",
"variation",
"autoregressive",
"lstm",
"svm"
],
[
"adversaries",
"attacker",
"attack",
"attacks",
"adversarially",
"adversarial",
"threat",
"robustness",
"adversary",
"malicious"
],
[
"datasets",
"modelling",
"transformer",
"model",
"models",
"manifold",
"objects",
"dataset",
"transformers",
"modeling"
],
[
"differentially",
"dynamical",
"observational",
"laplacian",
"analytical",
"analytically",
"differential",
"gradient",
"imply",
"derive"
],
[
"private",
"privacy",
"identity",
"embedding",
"extracted",
"classified",
"publicly",
"hidden",
"reveal",
"pass"
],
[
"expressed",
"topological",
"grounded",
"manifold",
"expression",
"nonconvex",
"semantic",
"intrinsic",
"embeddings",
"dimensionality"
],
[
"dedicated",
"flow",
"cycles",
"designs",
"illustrate",
"flows",
"exploit",
"designing",
"subspace",
"loop"
],
[
"huge",
"explored",
"volume",
"dependencies",
"ask",
"exploring",
"walk",
"remote",
"expand",
"numerous"
],
[
"bottleneck",
"handling",
"benchmarks",
"containing",
"sensitivity",
"contain",
"hyperparameters",
"mcmc",
"operator",
"kl"
],
[
"mixed",
"mixture",
"encoding",
"flow",
"stochastic",
"encoded",
"diffusion",
"information",
"flows",
"scheme"
],
[
"energy",
"devices",
"sensor",
"activity",
"static",
"stationary",
"eeg",
"brain",
"behaviour",
"wearable"
],
[
"guarantee",
"guaranteed",
"solver",
"deal",
"compatible",
"solvers",
"solves",
"guarantees",
"solving",
"solve"
],
[
"transformer",
"filters",
"cifar10",
"degrade",
"mri",
"degradation",
"imaging",
"pixels",
"decay",
"filtering"
],
[
"knowledge",
"distillation",
"learning",
"learnt",
"learned",
"kernel",
"studying",
"learner",
"learn",
"learners"
],
[
"measure",
"measurement",
"survey",
"practical",
"measurements",
"viable",
"measured",
"baselines",
"samples",
"step"
],
[
"posterior",
"partition",
"scalable",
"parallel",
"partitioning",
"hierarchical",
"multiple",
"hyperparameter",
"scalability",
"multi"
],
[
"exploration",
"black",
"critic",
"before",
"year",
"scratch",
"regret",
"remove",
"previous",
"date"
],
[
"ml",
"industrial",
"machines",
"machine",
"computer",
"multi",
"music",
"memory",
"mining",
"ai"
],
[
"separation",
"discrete",
"screening",
"dynamics",
"distillation",
"discriminator",
"controller",
"denoising",
"convolution",
"filtering"
],
[
"automated",
"classifier",
"classifiers",
"svm",
"automatically",
"embedded",
"labeling",
"algorithmic",
"textit",
"label"
],
[
"perception",
"sampling",
"sensing",
"detect",
"detection",
"detecting",
"detected",
"representative",
"susceptible",
"selection"
],
[
"ensembles",
"outside",
"weather",
"cloud",
"forest",
"forecast",
"forecasting",
"forecasts",
"environment",
"evolution"
],
[
"collect",
"preserving",
"compressed",
"sparse",
"compress",
"preserve",
"collected",
"vast",
"capturing",
"heavy"
],
[
"preferences",
"popularity",
"user",
"recommender",
"suggested",
"suggest",
"recommendations",
"recommendation",
"personalized",
"influence"
],
[
"points",
"heuristic",
"heuristics",
"deployment",
"inference",
"infinite",
"computations",
"reasoning",
"inherent",
"contrary"
],
[
"sensors",
"ratio",
"sensor",
"exposure",
"shot",
"low",
"neighbor",
"impractical",
"neighbors",
"short"
],
[
"approximations",
"converge",
"boundary",
"converges",
"limits",
"biases",
"limit",
"boundaries",
"convergence",
"offers"
],
[
"fewer",
"generalizes",
"smaller",
"characterizing",
"generalizing",
"generalize",
"generalization",
"generalizability",
"asymptotically",
"enough"
],
[
"satisfies",
"straightforward",
"softmax",
"minimax",
"posterior",
"corpus",
"fields",
"taxonomy",
"field",
"focus"
],
[
"rigorously",
"cell",
"granularity",
"coarse",
"noisy",
"rigorous",
"exists",
"presence",
"queries",
"cells"
],
[
"modes",
"multi",
"concrete",
"boxes",
"here",
"box",
"structures",
"dimensions",
"diverse",
"description"
],
[
"entropy",
"streaming",
"compression",
"storage",
"relevance",
"sequence",
"storing",
"vital",
"records",
"preserving"
],
[
"mcmc",
"multimodal",
"linear",
"multiple",
"nonlinear",
"minimax",
"mismatch",
"linearly",
"ml",
"resnet"
],
[
"randomized",
"random",
"diffusion",
"kernels",
"randomly",
"stochastic",
"kernel",
"svm",
"manifold",
"gaussian"
],
[
"biological",
"mechanism",
"genetic",
"molecular",
"protein",
"biology",
"interactions",
"communities",
"beneficial",
"cifar10"
],
[
"modeled",
"representation",
"environments",
"models",
"representations",
"modelling",
"modeling",
"accessible",
"represented",
"model"
],
[
"count",
"compatible",
"variant",
"markov",
"verified",
"configuration",
"adopted",
"categorical",
"variants",
"applicability"
],
[
"aspects",
"explainable",
"interpret",
"interpretability",
"interpretation",
"interpretable",
"explanations",
"explaining",
"interpreted",
"explanation"
],
[
"gan",
"critic",
"nn",
"tuning",
"gnn",
"gnns",
"dnn",
"dnns",
"test",
"agnostic"
],
[
"greedy",
"regret",
"bandits",
"bandit",
"exploiting",
"exploration",
"uncertain",
"exploits",
"naive",
"strategies"
]
] | 4.372668 | intfloat/e5-large-v2 | 0.902 | -0.360292 | 0.181611 | 0.909664 |
BBC News | GMM | 42 | 10 | [
[
"oil",
"economic",
"state",
"india",
"president",
"country",
"countries",
"economy",
"indian",
"aids"
],
[
"actor",
"films",
"film",
"oscar",
"awards",
"comedy",
"award",
"actress",
"best",
"movie"
],
[
"election",
"blair",
"labour",
"howard",
"party",
"tory",
"tories",
"secretary",
"minister",
"leader"
],
[
"shares",
"growth",
"stock",
"prices",
"sales",
"bank",
"yukos",
"deutsche",
"market",
"company"
],
[
"nations",
"england",
"robinson",
"rugby",
"wales",
"six",
"coach",
"ireland",
"half",
"side"
],
[
"microsoft",
"security",
"search",
"software",
"users",
"site",
"virus",
"internet",
"web",
"sites"
],
[
"arsenal",
"liverpool",
"manager",
"league",
"football",
"united",
"chelsea",
"club",
"goal",
"cup"
],
[
"album",
"band",
"song",
"singer",
"music",
"chart",
"show",
"songs",
"rock",
"pop"
],
[
"her",
"olympic",
"indoor",
"final",
"champion",
"race",
"title",
"championships",
"athens",
"tennis"
],
[
"sony",
"tv",
"phone",
"broadband",
"mobile",
"phones",
"digital",
"video",
"technology",
"mobiles"
]
] | 4.292319 | intfloat/e5-large-v2 | 1 | 0.096087 | 0.245972 | 0.925518 |
BBC News | GMM | 42 | 20 | [
[
"india",
"countries",
"aids",
"africa",
"straw",
"foreign",
"eu",
"lisbon",
"poverty",
"aid"
],
[
"awards",
"film",
"films",
"director",
"oscar",
"actor",
"actress",
"aviator",
"best",
"award"
],
[
"rights",
"lord",
"law",
"lords",
"police",
"mps",
"human",
"bill",
"minimum",
"committee"
],
[
"euro",
"dollar",
"president",
"budget",
"currency",
"deficit",
"bush",
"federal",
"prices",
"economic"
],
[
"nations",
"england",
"robinson",
"ireland",
"rugby",
"squad",
"lions",
"six",
"coach",
"wales"
],
[
"search",
"users",
"virus",
"google",
"microsoft",
"security",
"software",
"site",
"web",
"computer"
],
[
"city",
"madrid",
"liverpool",
"stadium",
"club",
"contract",
"arsenal",
"football",
"smith",
"manager"
],
[
"show",
"concert",
"singer",
"rock",
"band",
"cent",
"mtv",
"star",
"tickets",
"50"
],
[
"tennis",
"henman",
"seed",
"australian",
"open",
"roddick",
"round",
"match",
"federer",
"champion"
],
[
"digital",
"technology",
"sony",
"mobile",
"phones",
"gadget",
"video",
"gadgets",
"3g",
"games"
],
[
"penalty",
"gara",
"half",
"lewsey",
"ireland",
"england",
"try",
"irish",
"side",
"tries"
],
[
"minutes",
"goals",
"liverpool",
"goal",
"ball",
"united",
"striker",
"shot",
"celtic",
"scored"
],
[
"race",
"indoor",
"olympic",
"athens",
"holmes",
"championships",
"gold",
"champion",
"60m",
"athletics"
],
[
"online",
"broadband",
"bt",
"net",
"peer",
"service",
"internet",
"users",
"file",
"phone"
],
[
"growth",
"china",
"economy",
"yen",
"japan",
"economic",
"2004",
"trade",
"deficit",
"exports"
],
[
"blair",
"labour",
"party",
"tory",
"election",
"howard",
"brown",
"tories",
"prime",
"leader"
],
[
"chelsea",
"mourinho",
"league",
"champions",
"arsenal",
"barcelona",
"cup",
"fa",
"jose",
"real"
],
[
"shares",
"yukos",
"company",
"oil",
"firm",
"deutsche",
"gazprom",
"russian",
"deal",
"stock"
],
[
"song",
"album",
"award",
"awards",
"music",
"stone",
"best",
"robbie",
"band",
"pop"
],
[
"growth",
"rates",
"rise",
"sales",
"economy",
"rate",
"figures",
"bank",
"prices",
"interest"
]
] | 7.317652 | intfloat/e5-large-v2 | 0.925 | 0.0748 | 0.199772 | 0.93581 |
BBC News | GMM | 42 | 30 | [
[
"lisbon",
"economy",
"japan",
"economic",
"eu",
"european",
"targets",
"population",
"europe",
"gross"
],
[
"movie",
"box",
"actor",
"film",
"actress",
"aviator",
"oscar",
"oscars",
"films",
"hollywood"
],
[
"lord",
"rights",
"police",
"human",
"law",
"lords",
"terror",
"hunting",
"trial",
"bill"
],
[
"sites",
"virus",
"attacks",
"spam",
"traffic",
"microsoft",
"software",
"security",
"mail",
"websites"
],
[
"football",
"spanish",
"manager",
"smith",
"scottish",
"scotland",
"celtic",
"brazil",
"spain",
"players"
],
[
"tv",
"bt",
"digital",
"phones",
"phone",
"net",
"mobiles",
"mobile",
"services",
"broadband"
],
[
"roddick",
"match",
"federer",
"beat",
"seed",
"open",
"australian",
"tennis",
"champion",
"final"
],
[
"local",
"minimum",
"workers",
"age",
"council",
"pension",
"pay",
"unions",
"pensions",
"tax"
],
[
"penalty",
"half",
"goal",
"ball",
"ireland",
"try",
"gara",
"minutes",
"side",
"referee"
],
[
"tsunami",
"poverty",
"india",
"aid",
"aids",
"countries",
"debt",
"foreign",
"africa",
"finance"
],
[
"prime",
"blair",
"chancellor",
"howard",
"tory",
"election",
"brown",
"labour",
"leader",
"party"
],
[
"growth",
"dollar",
"economy",
"economic",
"exports",
"deficit",
"spending",
"bush",
"euro",
"budget"
],
[
"mac",
"mini",
"dvd",
"games",
"gadget",
"sony",
"gaming",
"nintendo",
"apple",
"gamers"
],
[
"relay",
"gold",
"lewis",
"athletics",
"olympic",
"medal",
"athens",
"collins",
"francis",
"holmes"
],
[
"book",
"films",
"award",
"best",
"awards",
"director",
"prize",
"actor",
"film",
"festival"
],
[
"athens",
"test",
"doping",
"drugs",
"banned",
"greek",
"athletes",
"collins",
"tests",
"olympics"
],
[
"interest",
"housing",
"prices",
"rates",
"slowdown",
"bank",
"lending",
"mortgage",
"house",
"figures"
],
[
"euros",
"company",
"profit",
"shares",
"profits",
"sales",
"firm",
"car",
"plant",
"executive"
],
[
"mps",
"mp",
"party",
"parties",
"labour",
"vote",
"election",
"eu",
"silk",
"straw"
],
[
"coach",
"wales",
"rugby",
"ireland",
"lions",
"nations",
"robinson",
"england",
"six",
"squad"
],
[
"league",
"club",
"chelsea",
"mourinho",
"united",
"liverpool",
"arsenal",
"champions",
"manager",
"cup"
],
[
"deutsche",
"stock",
"takeover",
"bid",
"exchange",
"shareholders",
"shares",
"offer",
"london",
"talks"
],
[
"hip",
"piracy",
"files",
"file",
"software",
"music",
"hop",
"peer",
"sharing",
"users"
],
[
"race",
"olympic",
"briton",
"000m",
"cross",
"champion",
"minutes",
"final",
"finish",
"bronze"
],
[
"sec",
"court",
"gazprom",
"oil",
"company",
"bankruptcy",
"russian",
"russia",
"yukos",
"firm"
],
[
"cent",
"singer",
"lee",
"50",
"her",
"star",
"jackson",
"man",
"mtv",
"show"
],
[
"race",
"jump",
"european",
"madrid",
"indoor",
"60m",
"championships",
"holmes",
"seconds",
"gardener"
],
[
"album",
"song",
"chart",
"music",
"band",
"rock",
"stone",
"robbie",
"pop",
"singer"
],
[
"technology",
"digital",
"3g",
"project",
"computer",
"mobile",
"users",
"video",
"radio",
"content"
],
[
"henman",
"australian",
"tennis",
"cup",
"event",
"injury",
"davis",
"round",
"open",
"wimbledon"
]
] | 10.078234 | intfloat/e5-large-v2 | 0.89 | 0.063133 | 0.201363 | 0.935121 |
BBC News | GMM | 42 | 40 | [
[
"lisbon",
"eu",
"population",
"european",
"economic",
"japan",
"economy",
"gross",
"targets",
"europe"
],
[
"festival",
"hollywood",
"actor",
"film",
"musical",
"films",
"theatre",
"star",
"oscar",
"ceremony"
],
[
"terror",
"human",
"lord",
"trial",
"rights",
"lords",
"suspects",
"clarke",
"terrorism",
"without"
],
[
"security",
"microsoft",
"virus",
"attacks",
"mail",
"spam",
"site",
"sites",
"windows",
"traffic"
],
[
"scotland",
"scottish",
"spanish",
"manager",
"smith",
"celtic",
"football",
"spain",
"assistant",
"cup"
],
[
"broadband",
"net",
"search",
"online",
"spam",
"bt",
"report",
"web",
"internet",
"users"
],
[
"beat",
"davis",
"roddick",
"spain",
"cup",
"match",
"my",
"singles",
"carlos",
"victory"
],
[
"minimum",
"workers",
"unions",
"pensions",
"local",
"pay",
"increase",
"pension",
"council",
"union"
],
[
"penalty",
"half",
"goal",
"ball",
"ireland",
"try",
"gara",
"minutes",
"side",
"referee"
],
[
"countries",
"tsunami",
"india",
"aids",
"aid",
"poverty",
"debt",
"argentina",
"africa",
"disaster"
],
[
"howard",
"election",
"blair",
"labour",
"brown",
"tory",
"prime",
"chancellor",
"party",
"tories"
],
[
"deficit",
"economic",
"budget",
"euro",
"dollar",
"exports",
"bush",
"economy",
"growth",
"spending"
],
[
"mac",
"nintendo",
"sony",
"gadget",
"apple",
"mini",
"machine",
"gaming",
"gadgets",
"ipod"
],
[
"gold",
"athletics",
"francis",
"athens",
"relay",
"olympic",
"lewis",
"medal",
"holmes",
"collins"
],
[
"film",
"dutch",
"prize",
"award",
"festival",
"yen",
"japan",
"films",
"director",
"la"
],
[
"doping",
"drugs",
"greek",
"banned",
"athletes",
"tests",
"collins",
"test",
"athens",
"olympics"
],
[
"interest",
"housing",
"prices",
"rates",
"slowdown",
"bank",
"lending",
"mortgage",
"house",
"figures"
],
[
"india",
"deal",
"firm",
"oil",
"company",
"car",
"china",
"production",
"plant",
"profits"
],
[
"silk",
"party",
"vote",
"parties",
"eu",
"students",
"straw",
"labour",
"election",
"lib"
],
[
"rugby",
"lions",
"ireland",
"six",
"wales",
"nations",
"coach",
"robinson",
"squad",
"england"
],
[
"league",
"cup",
"arsenal",
"barcelona",
"champions",
"chelsea",
"mourinho",
"united",
"uefa",
"ferguson"
],
[
"deutsche",
"shareholders",
"bid",
"shares",
"exchange",
"london",
"takeover",
"stock",
"offer",
"talks"
],
[
"software",
"peer",
"file",
"files",
"sharing",
"piracy",
"users",
"hop",
"music",
"hip"
],
[
"race",
"olympic",
"briton",
"000m",
"cross",
"champion",
"minutes",
"final",
"finish",
"bronze"
],
[
"russian",
"russia",
"bankruptcy",
"putin",
"oil",
"gazprom",
"yukos",
"khodorkovsky",
"auction",
"gas"
],
[
"jackson",
"star",
"man",
"singer",
"her",
"lee",
"spider",
"tour",
"she",
"recorded"
],
[
"indoor",
"60m",
"holmes",
"championships",
"jump",
"olympic",
"madrid",
"gardener",
"race",
"european"
],
[
"song",
"rock",
"album",
"stone",
"band",
"robbie",
"music",
"pop",
"chart",
"best"
],
[
"cell",
"technology",
"ibm",
"computing",
"computer",
"project",
"linux",
"chip",
"used",
"use"
],
[
"champion",
"her",
"she",
"wimbledon",
"open",
"injury",
"women",
"event",
"sydney",
"australian"
],
[
"mobile",
"phones",
"phone",
"3g",
"mobiles",
"technology",
"service",
"camera",
"networks",
"calls"
],
[
"open",
"roddick",
"australian",
"seed",
"federer",
"tennis",
"round",
"beat",
"match",
"champion"
],
[
"comedy",
"awards",
"actress",
"oscar",
"best",
"film",
"award",
"aviator",
"actor",
"nominated"
],
[
"definition",
"content",
"dvd",
"tv",
"digital",
"format",
"technology",
"programmes",
"dvds",
"mtv"
],
[
"rapper",
"50",
"brother",
"cent",
"rap",
"murder",
"celebrity",
"album",
"star",
"wilson"
],
[
"children",
"bill",
"hunting",
"committee",
"hunt",
"law",
"police",
"ban",
"mps",
"parliament"
],
[
"tickets",
"stage",
"fans",
"park",
"festival",
"singer",
"musical",
"band",
"album",
"concert"
],
[
"tim",
"davis",
"greg",
"cup",
"henman",
"murray",
"tennis",
"tie",
"match",
"final"
],
[
"club",
"steven",
"liverpool",
"arsenal",
"city",
"madrid",
"stadium",
"contract",
"manager",
"premiership"
],
[
"company",
"sec",
"airline",
"financial",
"profits",
"shares",
"firm",
"stock",
"firms",
"airlines"
]
] | 13.398251 | intfloat/e5-large-v2 | 0.84 | 0.046662 | 0.199381 | 0.933878 |
BBC News | GMM | 42 | 50 | [
[
"eu",
"targets",
"airlines",
"economic",
"india",
"japan",
"lisbon",
"air",
"european",
"economy"
],
[
"show",
"hollywood",
"ceremony",
"awards",
"oscar",
"star",
"films",
"festival",
"sky",
"film"
],
[
"suspects",
"clarke",
"rights",
"trial",
"human",
"terror",
"law",
"lord",
"lords",
"police"
],
[
"microsoft",
"attacks",
"spam",
"mail",
"sites",
"virus",
"security",
"site",
"traffic",
"windows"
],
[
"scotland",
"scottish",
"spanish",
"manager",
"smith",
"celtic",
"football",
"spain",
"assistant",
"cup"
],
[
"broadband",
"net",
"search",
"online",
"spam",
"bt",
"report",
"web",
"internet",
"users"
],
[
"beat",
"davis",
"roddick",
"spain",
"cup",
"match",
"my",
"singles",
"carlos",
"victory"
],
[
"council",
"pay",
"minimum",
"students",
"local",
"tax",
"lib",
"age",
"increase",
"jobs"
],
[
"goal",
"minutes",
"half",
"penalty",
"ball",
"try",
"gara",
"ireland",
"side",
"lead"
],
[
"tsunami",
"countries",
"aid",
"aids",
"poverty",
"india",
"disaster",
"finance",
"africa",
"brown"
],
[
"election",
"labour",
"brown",
"blair",
"howard",
"prime",
"chancellor",
"party",
"tory",
"minister"
],
[
"economy",
"exports",
"economic",
"growth",
"rate",
"quarter",
"unemployment",
"2004",
"china",
"domestic"
],
[
"apple",
"nintendo",
"sony",
"mini",
"ipod",
"gadgets",
"machine",
"mac",
"gadget",
"psp"
],
[
"relay",
"gold",
"lewis",
"athletics",
"olympic",
"medal",
"athens",
"collins",
"francis",
"holmes"
],
[
"director",
"festival",
"award",
"winner",
"death",
"cinema",
"van",
"dutch",
"film",
"regime"
],
[
"doping",
"drugs",
"greek",
"banned",
"athletes",
"tests",
"collins",
"test",
"athens",
"olympics"
],
[
"interest",
"housing",
"prices",
"rates",
"slowdown",
"bank",
"lending",
"mortgage",
"house",
"figures"
],
[
"shares",
"firm",
"euros",
"profits",
"company",
"oil",
"sales",
"heineken",
"plant",
"western"
],
[
"straw",
"party",
"silk",
"vote",
"eu",
"labour",
"election",
"parties",
"voting",
"id"
],
[
"robinson",
"england",
"rugby",
"lions",
"ireland",
"coach",
"six",
"woodward",
"nations",
"squad"
],
[
"league",
"champions",
"mourinho",
"barcelona",
"chelsea",
"arsenal",
"united",
"uefa",
"ferguson",
"cup"
],
[
"deutsche",
"shareholders",
"bid",
"shares",
"exchange",
"london",
"takeover",
"stock",
"offer",
"talks"
],
[
"piracy",
"peer",
"sharing",
"file",
"files",
"software",
"users",
"search",
"hop",
"hip"
],
[
"race",
"olympic",
"briton",
"000m",
"cross",
"champion",
"minutes",
"final",
"finish",
"bronze"
],
[
"russian",
"russia",
"bankruptcy",
"putin",
"oil",
"gazprom",
"yukos",
"khodorkovsky",
"auction",
"gas"
],
[
"her",
"star",
"lee",
"jackson",
"singer",
"studio",
"recorded",
"tour",
"police",
"love"
],
[
"race",
"jump",
"european",
"madrid",
"indoor",
"60m",
"championships",
"holmes",
"seconds",
"gardener"
],
[
"song",
"music",
"pop",
"stone",
"album",
"band",
"robbie",
"rock",
"chart",
"single"
],
[
"computer",
"use",
"project",
"linux",
"ibm",
"technology",
"trusted",
"using",
"used",
"students"
],
[
"women",
"australian",
"sydney",
"champion",
"open",
"injury",
"her",
"event",
"wimbledon",
"melbourne"
],
[
"mobile",
"phones",
"phone",
"3g",
"mobiles",
"technology",
"service",
"camera",
"networks",
"calls"
],
[
"open",
"roddick",
"australian",
"seed",
"federer",
"tennis",
"round",
"beat",
"match",
"champion"
],
[
"best",
"aviator",
"drake",
"supporting",
"actress",
"vera",
"nominated",
"staunton",
"awards",
"oscars"
],
[
"content",
"definition",
"tv",
"dvd",
"format",
"digital",
"technology",
"devices",
"dvds",
"programmes"
],
[
"celebrity",
"cent",
"50",
"brother",
"murder",
"rap",
"star",
"rapper",
"album",
"wilson"
],
[
"police",
"child",
"committee",
"parents",
"parliament",
"children",
"mps",
"school",
"bill",
"secretary"
],
[
"concert",
"festival",
"band",
"tickets",
"park",
"stage",
"sold",
"fans",
"musical",
"glasgow"
],
[
"cup",
"davis",
"henman",
"greg",
"tim",
"murray",
"tennis",
"tie",
"match",
"seed"
],
[
"manager",
"club",
"steven",
"liverpool",
"arsenal",
"stadium",
"contract",
"madrid",
"city",
"premiership"
],
[
"profits",
"sec",
"company",
"shares",
"firm",
"stock",
"china",
"financial",
"airline",
"airlines"
],
[
"box",
"dvd",
"movie",
"7m",
"film",
"films",
"office",
"comedy",
"oscar",
"starring"
],
[
"budget",
"bush",
"congress",
"deficit",
"spending",
"federal",
"administration",
"security",
"cuts",
"trillion"
],
[
"workers",
"unions",
"pension",
"correspondent",
"union",
"sector",
"choice",
"prescott",
"plant",
"800"
],
[
"halo",
"xbox",
"game",
"sequel",
"games",
"microsoft",
"san",
"titles",
"gamers",
"playstation"
],
[
"euro",
"deficit",
"dollar",
"currency",
"bush",
"exports",
"oil",
"trade",
"imports",
"prices"
],
[
"president",
"water",
"country",
"argentina",
"brazil",
"fuel",
"oil",
"state",
"economic",
"currency"
],
[
"law",
"police",
"act",
"hunting",
"hunt",
"borders",
"supporters",
"ban",
"alliance",
"offence"
],
[
"prize",
"theatre",
"awards",
"award",
"her",
"book",
"film",
"actor",
"comedy",
"won"
],
[
"wales",
"rugby",
"france",
"nations",
"coach",
"edinburgh",
"welsh",
"cardiff",
"williams",
"six"
],
[
"chip",
"sony",
"intel",
"playstation",
"cell",
"power",
"light",
"technology",
"ibm",
"computing"
]
] | 16.230079 | intfloat/e5-large-v2 | 0.82 | 0.03737 | 0.191063 | 0.933611 |
BBC News | KeyNMF | 42 | 10 | [
[
"in",
"the",
"and",
"us",
"uk",
"new",
"was",
"has",
"music",
"show"
],
[
"labour",
"government",
"election",
"would",
"voters",
"campaign",
"uk",
"conservatives",
"tories",
"and"
],
[
"deal",
"investors",
"shares",
"and",
"companies",
"business",
"company",
"financial",
"firms",
"buy"
],
[
"broadband",
"phone",
"users",
"computers",
"software",
"technology",
"computer",
"internet",
"phones",
"devices"
],
[
"actress",
"nominated",
"actor",
"oscar",
"award",
"movie",
"awards",
"film",
"films",
"prize"
],
[
"from",
"security",
"information",
"new",
"told",
"news",
"japan",
"and",
"2004",
"involved"
],
[
"dollar",
"economic",
"spending",
"prices",
"rates",
"economist",
"us",
"economy",
"inflation",
"growth"
],
[
"some",
"has",
"also",
"because",
"according",
"us",
"sony",
"which",
"said",
"and"
],
[
"could",
"and",
"law",
"no",
"but",
"although",
"would",
"said",
"asked",
"not"
],
[
"england",
"had",
"has",
"wales",
"coach",
"players",
"injury",
"rugby",
"ireland",
"and"
]
] | 7.374024 | intfloat/e5-large-v2 | 0.86 | 0.034378 | 0.240613 | 0.779422 |
BBC News | KeyNMF | 42 | 20 | [
[
"london",
"or",
"and",
"against",
"like",
"added",
"new",
"now",
"however",
"alleged"
],
[
"voters",
"labour",
"issues",
"vote",
"campaign",
"election",
"general",
"tories",
"conservatives",
"political"
],
[
"exports",
"prices",
"economist",
"spending",
"economic",
"growth",
"inflation",
"economy",
"rates",
"dollar"
],
[
"mobile",
"gadgets",
"phone",
"technology",
"technologies",
"phones",
"mobiles",
"devices",
"nokia",
"download"
],
[
"films",
"nominated",
"film",
"award",
"awards",
"oscar",
"actor",
"actress",
"movie",
"oscars"
],
[
"news",
"told",
"new",
"from",
"2004",
"information",
"japan",
"involved",
"see",
"production"
],
[
"but",
"although",
"because",
"further",
"no",
"asked",
"not",
"law",
"however",
"says"
],
[
"said",
"according",
"some",
"because",
"including",
"recently",
"which",
"also",
"german",
"told"
],
[
"would",
"could",
"told",
"should",
"suggested",
"think",
"asked",
"like",
"concerns",
"costs"
],
[
"ireland",
"players",
"france",
"rugby",
"coach",
"wales",
"england",
"six",
"nations",
"scotland"
],
[
"uk",
"british",
"european",
"london",
"britain",
"according",
"europe",
"broadband",
"housing",
"countries"
],
[
"websites",
"computer",
"security",
"it",
"software",
"users",
"computers",
"internet",
"windows",
"microsoft"
],
[
"shares",
"companies",
"business",
"investors",
"financial",
"company",
"deal",
"firms",
"buy",
"investment"
],
[
"new",
"show",
"music",
"the",
"in",
"by",
"singer",
"was",
"to",
"band"
],
[
"tories",
"budget",
"spending",
"ministers",
"plans",
"taxes",
"tax",
"government",
"people",
"pensions"
],
[
"have",
"in",
"has",
"injury",
"with",
"his",
"he",
"added",
"premiership",
"out"
],
[
"olympic",
"winning",
"indoor",
"champion",
"championships",
"olympics",
"international",
"in",
"athletics",
"britain"
],
[
"then",
"said",
"was",
"players",
"had",
"his",
"played",
"looked",
"victory",
"match"
],
[
"us",
"according",
"dollar",
"america",
"filed",
"legal",
"firms",
"companies",
"deficit",
"federal"
],
[
"game",
"games",
"sony",
"gaming",
"sales",
"consoles",
"console",
"playstation",
"xbox",
"handheld"
]
] | 7.087672 | intfloat/e5-large-v2 | 0.88 | 0.009296 | 0.239423 | 0.82082 |
BBC News | KeyNMF | 42 | 30 | [
[
"but",
"because",
"although",
"however",
"asked",
"some",
"further",
"no",
"police",
"not"
],
[
"campaign",
"tories",
"election",
"labour",
"voters",
"vote",
"issues",
"political",
"conservatives",
"general"
],
[
"dollar",
"economy",
"exports",
"growth",
"economic",
"spending",
"inflation",
"rates",
"economist",
"analysts"
],
[
"electronics",
"technologies",
"digital",
"broadband",
"technology",
"devices",
"gadgets",
"images",
"tech",
"cameras"
],
[
"nominated",
"awards",
"actress",
"prize",
"oscar",
"award",
"oscars",
"nominations",
"winning",
"actor"
],
[
"new",
"told",
"from",
"information",
"news",
"japan",
"see",
"involved",
"2004",
"her"
],
[
"prices",
"price",
"sales",
"rates",
"housing",
"rate",
"consumers",
"december",
"figures",
"mortgage"
],
[
"recently",
"some",
"also",
"said",
"which",
"including",
"because",
"next",
"german",
"continued"
],
[
"film",
"starring",
"the",
"cinema",
"films",
"movie",
"movies",
"actor",
"comedy",
"hollywood"
],
[
"could",
"talks",
"transport",
"people",
"costs",
"cost",
"unions",
"speculation",
"research",
"told"
],
[
"britain",
"uk",
"british",
"london",
"europe",
"countries",
"european",
"foreign",
"immigration",
"eu"
],
[
"windows",
"computers",
"users",
"computer",
"security",
"software",
"pc",
"it",
"microsoft",
"viruses"
],
[
"financial",
"shares",
"business",
"stock",
"shareholders",
"investors",
"company",
"buy",
"deal",
"takeover"
],
[
"to",
"january",
"international",
"india",
"win",
"indian",
"her",
"the",
"in",
"she"
],
[
"russian",
"firms",
"companies",
"filed",
"company",
"business",
"russia",
"state",
"yukos",
"corporate"
],
[
"has",
"injury",
"his",
"have",
"with",
"he",
"added",
"been",
"coach",
"had"
],
[
"athletics",
"champion",
"championships",
"indoor",
"olympic",
"winning",
"olympics",
"britain",
"sport",
"athletes"
],
[
"victory",
"then",
"said",
"match",
"played",
"was",
"looked",
"his",
"had",
"points"
],
[
"trade",
"america",
"us",
"sales",
"deficit",
"dollar",
"currency",
"deficits",
"federal",
"american"
],
[
"much",
"news",
"2004",
"christmas",
"according",
"china",
"chinese",
"2003",
"newspaper",
"2005"
],
[
"and",
"or",
"like",
"against",
"alleged",
"former",
"added",
"meanwhile",
"however",
"now"
],
[
"new",
"the",
"band",
"singer",
"music",
"by",
"songs",
"album",
"latest",
"who"
],
[
"gaming",
"sony",
"game",
"games",
"console",
"playstation",
"nintendo",
"xbox",
"consoles",
"handheld"
],
[
"phones",
"handsets",
"mobiles",
"mobile",
"nokia",
"phone",
"wireless",
"3g",
"handset",
"calls"
],
[
"budget",
"plans",
"government",
"tories",
"tax",
"spending",
"taxes",
"ministers",
"people",
"pensions"
],
[
"media",
"viewers",
"bbc",
"tv",
"show",
"television",
"shows",
"broadcast",
"audience",
"comedy"
],
[
"wales",
"ireland",
"rugby",
"france",
"england",
"players",
"coach",
"nations",
"six",
"scotland"
],
[
"says",
"should",
"told",
"said",
"think",
"legal",
"asked",
"parliament",
"law",
"would"
],
[
"arsenal",
"players",
"player",
"united",
"football",
"champions",
"premiership",
"chelsea",
"liverpool",
"manager"
],
[
"google",
"users",
"internet",
"broadband",
"web",
"website",
"websites",
"search",
"like",
"information"
]
] | 7.620865 | intfloat/e5-large-v2 | 0.883333 | 0.019782 | 0.249355 | 0.839831 |
BBC News | KeyNMF | 42 | 40 | [
[
"because",
"although",
"but",
"however",
"though",
"not",
"further",
"no",
"why",
"some"
],
[
"election",
"labour",
"conservatives",
"vote",
"voters",
"tories",
"political",
"campaign",
"general",
"parties"
],
[
"dollar",
"exports",
"prices",
"currency",
"china",
"economy",
"trade",
"economic",
"euro",
"trading"
],
[
"web",
"websites",
"users",
"website",
"google",
"internet",
"search",
"sites",
"using",
"information"
],
[
"awards",
"award",
"winner",
"nominated",
"winners",
"voted",
"best",
"prize",
"won",
"categories"
],
[
"from",
"see",
"information",
"recent",
"told",
"new",
"decision",
"her",
"news",
"2004"
],
[
"economic",
"economist",
"growth",
"rates",
"economy",
"spending",
"prices",
"consumer",
"inflation",
"rate"
],
[
"also",
"recently",
"some",
"including",
"because",
"which",
"firms",
"continued",
"next",
"general"
],
[
"actor",
"nominations",
"oscar",
"aviator",
"nominated",
"oscars",
"nomination",
"awards",
"award",
"actress"
],
[
"players",
"rugby",
"wales",
"ireland",
"france",
"england",
"coach",
"nations",
"six",
"scotland"
],
[
"british",
"housing",
"london",
"uk",
"britain",
"immigration",
"foreign",
"mortgage",
"price",
"house"
],
[
"software",
"windows",
"computer",
"it",
"security",
"computers",
"pc",
"users",
"viruses",
"microsoft"
],
[
"deal",
"shares",
"investors",
"financial",
"buy",
"business",
"shareholders",
"stock",
"offer",
"takeover"
],
[
"in",
"international",
"indian",
"to",
"january",
"india",
"her",
"she",
"the",
"win"
],
[
"technology",
"devices",
"digital",
"electronics",
"technologies",
"gadgets",
"images",
"tech",
"cameras",
"consumers"
],
[
"his",
"with",
"added",
"had",
"have",
"has",
"injury",
"he",
"been",
"out"
],
[
"winning",
"indoor",
"champion",
"britain",
"championships",
"olympic",
"seconds",
"indoors",
"birmingham",
"hurdles"
],
[
"says",
"told",
"london",
"politics",
"said",
"say",
"think",
"would",
"news",
"jewish"
],
[
"spending",
"us",
"american",
"federal",
"america",
"sales",
"firm",
"securities",
"usa",
"americans"
],
[
"speeds",
"internet",
"now",
"bt",
"phone",
"customers",
"net",
"speed",
"broadband",
"telecoms"
],
[
"parliament",
"law",
"against",
"laws",
"would",
"legal",
"rights",
"legislation",
"court",
"whether"
],
[
"the",
"music",
"singer",
"album",
"new",
"by",
"band",
"songs",
"his",
"next"
],
[
"victory",
"match",
"was",
"his",
"played",
"looked",
"then",
"had",
"points",
"final"
],
[
"phone",
"mobile",
"handsets",
"mobiles",
"nokia",
"phones",
"3g",
"wireless",
"handset",
"using"
],
[
"games",
"game",
"gaming",
"sony",
"console",
"playstation",
"handheld",
"consoles",
"xbox",
"nintendo"
],
[
"television",
"viewers",
"tv",
"show",
"bbc",
"media",
"shows",
"audience",
"broadcast",
"comedy"
],
[
"film",
"films",
"movie",
"cinema",
"movies",
"starring",
"the",
"comedy",
"hollywood",
"critics"
],
[
"business",
"firms",
"company",
"investors",
"corporate",
"companies",
"businesses",
"industry",
"industries",
"prosecutors"
],
[
"could",
"should",
"would",
"suggested",
"talks",
"told",
"concerns",
"speculation",
"cards",
"concern"
],
[
"premiership",
"players",
"chelsea",
"player",
"football",
"united",
"arsenal",
"champions",
"goals",
"manager"
],
[
"december",
"2004",
"sales",
"christmas",
"according",
"retailers",
"news",
"festive",
"prices",
"much"
],
[
"tax",
"budget",
"spending",
"government",
"tories",
"plans",
"taxes",
"ministers",
"pensions",
"people"
],
[
"germany",
"europe",
"european",
"eu",
"euro",
"says",
"german",
"euros",
"countries",
"france"
],
[
"russian",
"russia",
"bankruptcy",
"oil",
"yukos",
"gazprom",
"state",
"tax",
"filed",
"putin"
],
[
"olympic",
"athletes",
"sport",
"drugs",
"olympics",
"athletics",
"doping",
"sprinter",
"athens",
"trial"
],
[
"like",
"much",
"would",
"important",
"already",
"want",
"really",
"deal",
"match",
"music"
],
[
"passengers",
"cost",
"costs",
"prices",
"airline",
"airlines",
"travel",
"customers",
"price",
"carriers"
],
[
"asked",
"questions",
"poll",
"says",
"comments",
"question",
"issues",
"replied",
"no",
"report"
],
[
"and",
"or",
"recent",
"meanwhile",
"against",
"alleged",
"now",
"former",
"vehicles",
"house"
],
[
"nations",
"countries",
"tsunami",
"aid",
"world",
"africa",
"developing",
"international",
"poverty",
"finance"
]
] | 8.259519 | intfloat/e5-large-v2 | 0.8425 | 0.011677 | 0.248919 | 0.855483 |
BBC News | KeyNMF | 42 | 50 | [
[
"but",
"although",
"because",
"however",
"not",
"further",
"no",
"though",
"some",
"criticised"
],
[
"conservatives",
"tories",
"conservative",
"taxes",
"howard",
"tax",
"immigration",
"tory",
"labour",
"plans"
],
[
"exports",
"euro",
"currency",
"dollar",
"trade",
"china",
"prices",
"deficit",
"trading",
"deficits"
],
[
"google",
"website",
"users",
"internet",
"search",
"like",
"web",
"websites",
"sites",
"information"
],
[
"campaign",
"political",
"election",
"labour",
"voters",
"vote",
"general",
"issues",
"elections",
"parties"
],
[
"japan",
"told",
"march",
"recent",
"2004",
"see",
"already",
"from",
"information",
"new"
],
[
"economy",
"economic",
"growth",
"rates",
"spending",
"economist",
"consumer",
"inflation",
"forecast",
"jobs"
],
[
"which",
"recently",
"also",
"some",
"because",
"next",
"continued",
"rules",
"including",
"scottish"
],
[
"best",
"voted",
"won",
"winners",
"winner",
"prize",
"awards",
"award",
"nominated",
"categories"
],
[
"could",
"told",
"confident",
"interested",
"suggested",
"believes",
"speculation",
"concerns",
"potential",
"used"
],
[
"british",
"london",
"britain",
"rock",
"immigration",
"foreign",
"uk",
"housing",
"needed",
"says"
],
[
"it",
"pc",
"computers",
"software",
"computer",
"windows",
"users",
"pcs",
"apple",
"using"
],
[
"nominated",
"actor",
"actress",
"nominations",
"awards",
"award",
"oscar",
"oscars",
"nomination",
"aviator"
],
[
"january",
"in",
"indian",
"to",
"international",
"india",
"the",
"or",
"irish",
"are"
],
[
"victory",
"scored",
"played",
"then",
"points",
"defence",
"referee",
"looked",
"penalty",
"added"
],
[
"with",
"have",
"injury",
"his",
"has",
"added",
"he",
"been",
"contract",
"out"
],
[
"championships",
"olympic",
"indoor",
"winning",
"champion",
"international",
"britain",
"birmingham",
"indoors",
"seconds"
],
[
"win",
"her",
"match",
"was",
"tennis",
"she",
"playing",
"tournament",
"australian",
"final"
],
[
"us",
"american",
"america",
"federal",
"securities",
"sales",
"usa",
"firm",
"spending",
"americans"
],
[
"customers",
"internet",
"broadband",
"speeds",
"speed",
"telecoms",
"phone",
"bt",
"net",
"now"
],
[
"law",
"rights",
"legal",
"laws",
"courts",
"court",
"legislation",
"against",
"parliament",
"police"
],
[
"new",
"by",
"the",
"music",
"his",
"album",
"singer",
"band",
"songs",
"next"
],
[
"worked",
"died",
"former",
"then",
"was",
"his",
"did",
"had",
"later",
"have"
],
[
"phones",
"mobile",
"mobiles",
"handsets",
"handset",
"nokia",
"3g",
"phone",
"wireless",
"use"
],
[
"handheld",
"games",
"console",
"game",
"sony",
"xbox",
"gaming",
"consoles",
"nintendo",
"playstation"
],
[
"media",
"show",
"broadcast",
"viewers",
"comedy",
"bbc",
"audience",
"shows",
"tv",
"television"
],
[
"film",
"films",
"movies",
"cinema",
"starring",
"movie",
"the",
"hollywood",
"comedy",
"new"
],
[
"england",
"coach",
"six",
"france",
"rugby",
"wales",
"scotland",
"ireland",
"nations",
"injury"
],
[
"buy",
"shares",
"takeover",
"offer",
"deal",
"shareholders",
"stock",
"share",
"bought",
"buying"
],
[
"product",
"fees",
"against",
"now",
"like",
"clients",
"and",
"vehicles",
"can",
"know"
],
[
"2003",
"tax",
"much",
"china",
"christmas",
"according",
"2004",
"chinese",
"research",
"gdp"
],
[
"chelsea",
"liverpool",
"premiership",
"united",
"football",
"arsenal",
"champions",
"manager",
"mourinho",
"goals"
],
[
"people",
"pensions",
"workers",
"pension",
"unions",
"talks",
"transport",
"union",
"meeting",
"labour"
],
[
"yukos",
"state",
"oil",
"russia",
"bankruptcy",
"putin",
"filed",
"russian",
"gazprom",
"tax"
],
[
"his",
"want",
"sport",
"player",
"football",
"players",
"like",
"rugby",
"winning",
"sure"
],
[
"media",
"news",
"english",
"interested",
"radio",
"yet",
"much",
"about",
"spanish",
"reporters"
],
[
"technologies",
"technology",
"gadgets",
"devices",
"electronics",
"digital",
"images",
"tech",
"cameras",
"media"
],
[
"investment",
"investors",
"financial",
"analysts",
"profits",
"shares",
"securities",
"company",
"corporate",
"shareholder"
],
[
"christmas",
"retailers",
"december",
"housing",
"retail",
"prices",
"consumers",
"price",
"sales",
"figures"
],
[
"aid",
"countries",
"international",
"developing",
"poverty",
"tsunami",
"nations",
"world",
"africa",
"finance"
],
[
"costs",
"cost",
"airline",
"passengers",
"airlines",
"travel",
"euros",
"air",
"carriers",
"customers"
],
[
"doping",
"drugs",
"sport",
"athletics",
"trial",
"olympics",
"sprinter",
"olympic",
"athletes",
"tests"
],
[
"investors",
"companies",
"company",
"production",
"businesses",
"industries",
"industry",
"firms",
"states",
"shares"
],
[
"business",
"concerned",
"company",
"savings",
"research",
"money",
"fraud",
"investment",
"selling",
"developers"
],
[
"suggested",
"want",
"concern",
"would",
"parliament",
"should",
"think",
"debate",
"says",
"teachers"
],
[
"ministers",
"finance",
"concerns",
"plans",
"budget",
"money",
"government",
"spending",
"treasury",
"state"
],
[
"viruses",
"attacks",
"websites",
"mail",
"security",
"attack",
"virus",
"software",
"warned",
"users"
],
[
"like",
"says",
"think",
"jewish",
"say",
"london",
"told",
"politics",
"said",
"accused"
],
[
"euro",
"countries",
"europe",
"euros",
"european",
"says",
"eu",
"german",
"germany",
"europeans"
],
[
"asked",
"replied",
"questions",
"says",
"no",
"comments",
"question",
"comment",
"poll",
"report"
]
] | 9.966823 | intfloat/e5-large-v2 | 0.828 | -0.021573 | 0.2377 | 0.856805 |
BBC News | FASTopic | 42 | 10 | [
[
"star",
"best",
"album",
"director",
"band",
"awards",
"film",
"song",
"award",
"films"
],
[
"the",
"of",
"to",
"and",
"in",
"for",
"that",
"is",
"it",
"on"
],
[
"election",
"government",
"labour",
"party",
"blair",
"minister",
"plans",
"tax",
"brown",
"howard"
],
[
"controversial",
"hunt",
"ruling",
"watchdog",
"criminal",
"courts",
"cases",
"spokeswoman",
"guilty",
"justice"
],
[
"case",
"russia",
"bid",
"sale",
"deutsche",
"yukos",
"russian",
"talks",
"court",
"drugs"
],
[
"champion",
"open",
"women",
"race",
"her",
"title",
"old",
"final",
"olympic",
"mark"
],
[
"users",
"technology",
"digital",
"tv",
"video",
"software",
"mobile",
"net",
"phone",
"data"
],
[
"growth",
"economy",
"economic",
"rise",
"figures",
"prices",
"interest",
"rate",
"strong",
"sales"
],
[
"ireland",
"france",
"cup",
"side",
"match",
"wales",
"club",
"playing",
"united",
"england"
],
[
"financial",
"shares",
"firm",
"india",
"companies",
"oil",
"countries",
"firms",
"company",
"state"
]
] | 67.890411 | intfloat/e5-large-v2 | 1 | 0.027087 | 0.202059 | 0.886572 |
BBC News | FASTopic | 42 | 20 | [
[
"election",
"blair",
"howard",
"labour",
"brown",
"party",
"leader",
"lord",
"tory",
"prime"
],
[
"nintendo",
"movies",
"picture",
"films",
"format",
"moving",
"model",
"commercial",
"console",
"visitors"
],
[
"game",
"players",
"play",
"my",
"final",
"old",
"france",
"win",
"half",
"ireland"
],
[
"of",
"the",
"to",
"and",
"in",
"for",
"is",
"that",
"it",
"on"
],
[
"england",
"scotland",
"wales",
"rugby",
"club",
"coach",
"injury",
"united",
"league",
"liverpool"
],
[
"ruling",
"ban",
"trial",
"appeal",
"allowed",
"criminal",
"ruled",
"charge",
"held",
"illegal"
],
[
"indoor",
"jump",
"glasgow",
"cross",
"van",
"race",
"edinburgh",
"lewis",
"birmingham",
"chris"
],
[
"software",
"media",
"video",
"music",
"tv",
"digital",
"users",
"computer",
"technology",
"radio"
],
[
"ones",
"type",
"additional",
"aims",
"foundation",
"helps",
"california",
"page",
"engines",
"region"
],
[
"china",
"bank",
"economy",
"deal",
"economic",
"sales",
"growth",
"market",
"president",
"company"
],
[
"government",
"law",
"court",
"public",
"police",
"spokesman",
"bill",
"case",
"under",
"rights"
],
[
"services",
"phones",
"broadband",
"according",
"phone",
"mobile",
"research",
"sony",
"million",
"customers"
],
[
"dutch",
"became",
"brother",
"books",
"2003",
"aged",
"lawsuit",
"wife",
"friend",
"documentary"
],
[
"round",
"champion",
"australian",
"drugs",
"open",
"olympic",
"pair",
"tournament",
"matches",
"athens"
],
[
"russian",
"deutsche",
"indian",
"talks",
"bid",
"agency",
"unit",
"owned",
"shareholders",
"management"
],
[
"tsunami",
"1997",
"jack",
"disaster",
"payments",
"agenda",
"votes",
"funding",
"skills",
"spring"
],
[
"goals",
"drawn",
"mourinho",
"striker",
"steven",
"brazil",
"alex",
"fa",
"clubs",
"uefa"
],
[
"award",
"best",
"band",
"film",
"awards",
"star",
"song",
"album",
"director",
"box"
],
[
"january",
"air",
"federal",
"airline",
"airlines",
"continuing",
"yen",
"optimistic",
"grew",
"stronger"
],
[
"spanish",
"festival",
"chart",
"prince",
"singer",
"tour",
"woman",
"died",
"rap",
"cent"
]
] | 1,726.299443 | intfloat/e5-large-v2 | 1 | -0.151701 | 0.169501 | 0.907169 |
BBC News | FASTopic | 42 | 30 | [
[
"hoping",
"hopes",
"beginning",
"summit",
"near",
"factor",
"considering",
"elsewhere",
"progress",
"handling"
],
[
"flying",
"realise",
"straight",
"open",
"french",
"wild",
"gear",
"roger",
"fired",
"warm"
],
[
"owners",
"car",
"lines",
"increasingly",
"handle",
"helps",
"wave",
"tiny",
"launches",
"affordable"
],
[
"profits",
"profit",
"85",
"congress",
"earnings",
"2010",
"cars",
"blamed",
"400",
"estimates"
],
[
"such",
"music",
"services",
"data",
"tv",
"people",
"radio",
"many",
"says",
"those"
],
[
"manager",
"arsenal",
"tie",
"barcelona",
"mourinho",
"stadium",
"jose",
"federer",
"steven",
"phillips"
],
[
"half",
"game",
"players",
"england",
"final",
"team",
"wales",
"france",
"ireland",
"side"
],
[
"bid",
"india",
"stock",
"shares",
"german",
"deutsche",
"talks",
"investors",
"air",
"agency"
],
[
"consider",
"aware",
"prevent",
"initiative",
"spread",
"seeking",
"threats",
"hundreds",
"customer",
"criminals"
],
[
"connection",
"technical",
"works",
"malicious",
"downloaded",
"useful",
"broadcast",
"mtv",
"station",
"tools"
],
[
"sold",
"production",
"rival",
"russia",
"indian",
"sale",
"sell",
"russian",
"giant",
"selling"
],
[
"very",
"my",
"him",
"his",
"got",
"think",
"don",
"me",
"right",
"good"
],
[
"spring",
"duty",
"agenda",
"calling",
"reforms",
"funding",
"reduced",
"payments",
"transport",
"retirement"
],
[
"award",
"awards",
"director",
"film",
"films",
"box",
"prize",
"movie",
"actor",
"oscar"
],
[
"disaster",
"victims",
"jackson",
"killed",
"tsunami",
"died",
"described",
"relief",
"death",
"books"
],
[
"refused",
"alan",
"decision",
"situation",
"sunday",
"monday",
"insisted",
"failed",
"wednesday",
"absolutely"
],
[
"labour",
"brown",
"election",
"party",
"general",
"howard",
"blair",
"leader",
"lord",
"minister"
],
[
"australian",
"athens",
"race",
"compete",
"indoor",
"gold",
"titles",
"olympic",
"champion",
"championships"
],
[
"commit",
"join",
"session",
"arrested",
"secret",
"honest",
"supported",
"sensible",
"shall",
"representing"
],
[
"steve",
"fit",
"edinburgh",
"glasgow",
"brian",
"fantastic",
"simon",
"definitely",
"bath",
"thompson"
],
[
"budget",
"government",
"tax",
"state",
"foreign",
"countries",
"eu",
"national",
"case",
"commission"
],
[
"users",
"net",
"mobile",
"technology",
"digital",
"software",
"internet",
"computer",
"video",
"phone"
],
[
"charged",
"drugs",
"drug",
"suspended",
"dutch",
"doping",
"brother",
"banned",
"factory",
"incident"
],
[
"sales",
"china",
"president",
"growth",
"business",
"bank",
"economy",
"figures",
"market",
"2004"
],
[
"the",
"of",
"to",
"and",
"that",
"it",
"is",
"said",
"be",
"on"
],
[
"legal",
"system",
"free",
"offer",
"currently",
"executive",
"firms",
"children",
"based",
"companies"
],
[
"best",
"second",
"her",
"won",
"she",
"top",
"three",
"first",
"four",
"record"
],
[
"prince",
"band",
"pop",
"fans",
"singer",
"woman",
"album",
"star",
"musical",
"rock"
],
[
"japanese",
"overall",
"fell",
"index",
"recovery",
"domestic",
"japan",
"currency",
"mortgage",
"korea"
],
[
"ban",
"court",
"ruling",
"accused",
"judge",
"hunting",
"ex",
"comment",
"straw",
"inquiry"
]
] | 5,216.889036 | intfloat/e5-large-v2 | 1 | -0.159417 | 0.178834 | 0.900499 |
BBC News | S³_angular | 42 | 10 | [
[
"england",
"briton",
"britons",
"britain",
"british",
"tories",
"uk",
"westminster",
"gb",
"london"
],
[
"spammers",
"reported",
"knew",
"caught",
"voted",
"didn",
"those",
"stolen",
"responded",
"resulted"
],
[
"advertising",
"unions",
"payments",
"pensions",
"directors",
"consumption",
"disney",
"pension",
"bills",
"cap"
],
[
"issues",
"government",
"legislation",
"opposed",
"governments",
"policies",
"scrutiny",
"concerns",
"restrictions",
"immigration"
],
[
"agency",
"agencies",
"zealand",
"regulation",
"members",
"forms",
"regulators",
"regulator",
"assess",
"bath"
],
[
"talks",
"mp",
"commission",
"speaking",
"speech",
"meetings",
"public",
"talking",
"offices",
"interests"
],
[
"increases",
"inflation",
"growth",
"recession",
"households",
"incomes",
"increasing",
"forecasts",
"rates",
"economic"
],
[
"skills",
"backs",
"youth",
"shouldn",
"years",
"careful",
"children",
"programs",
"should",
"culture"
],
[
"mobiles",
"broadband",
"gadgets",
"technology",
"handsets",
"speeds",
"laptops",
"technologies",
"phones",
"handheld"
],
[
"perhaps",
"potential",
"possibly",
"potentially",
"midfielder",
"consultancy",
"could",
"coaching",
"manager",
"managing"
]
] | 3.254126 | intfloat/e5-large-v2 | 1 | -0.316866 | 0.225001 | 0.889878 |
BBC News | S³_angular | 42 | 20 | [
[
"resigned",
"hodgson",
"uk",
"mortgage",
"britons",
"westminster",
"whilst",
"rumours",
"fewer",
"gone"
],
[
"savings",
"budget",
"costs",
"budgets",
"saving",
"providers",
"finances",
"pensions",
"cost",
"spending"
],
[
"sharing",
"onto",
"divided",
"replacing",
"lay",
"skipper",
"threw",
"letwin",
"instead",
"presenter"
],
[
"result",
"compared",
"results",
"dominated",
"resulted",
"arrived",
"battled",
"ourselves",
"counterparts",
"attempts"
],
[
"athletics",
"hurdles",
"competing",
"olympics",
"running",
"sprinter",
"runners",
"qualifying",
"medallist",
"athletes"
],
[
"courts",
"constitutional",
"lawyers",
"waiting",
"wait",
"judge",
"whether",
"discussions",
"court",
"restrictions"
],
[
"british",
"championships",
"wimbledon",
"ultimately",
"negotiations",
"billion",
"britain",
"birmingham",
"luxury",
"20bn"
],
[
"court",
"roddick",
"matches",
"wimbledon",
"tennis",
"federer",
"courts",
"rankings",
"maria",
"tournament"
],
[
"itunes",
"analysts",
"executives",
"expectations",
"suggestions",
"networks",
"contenders",
"interviews",
"shouldn",
"critics"
],
[
"unlikely",
"forgotten",
"abandoned",
"scrapped",
"slowdown",
"uncertain",
"forget",
"uncertainty",
"doubts",
"slower"
],
[
"filing",
"attorney",
"prosecution",
"lawsuits",
"lawsuit",
"prosecutors",
"contracts",
"stevens",
"alleged",
"investigation"
],
[
"asset",
"investments",
"analyst",
"screening",
"investment",
"thinks",
"intelligence",
"bidding",
"secretary",
"henman"
],
[
"xbox",
"playstation",
"dvds",
"consoles",
"sony",
"nintendo",
"dvd",
"handheld",
"psp",
"anticipated"
],
[
"recording",
"mobile",
"initially",
"3g",
"download",
"terms",
"midfield",
"eventually",
"nokia",
"later"
],
[
"disappointed",
"worrying",
"criticised",
"unacceptable",
"worried",
"blamed",
"badly",
"disappointing",
"damaging",
"disappointment"
],
[
"holmes",
"country",
"legislation",
"countries",
"unveils",
"land",
"asia",
"canada",
"infrastructure",
"australia"
],
[
"tory",
"candidates",
"campaigning",
"election",
"conservative",
"blair",
"tories",
"candidate",
"campaign",
"elections"
],
[
"publicly",
"multimedia",
"maria",
"mother",
"amazon",
"imelda",
"nokia",
"markets",
"market",
"ago"
],
[
"scottish",
"scotland",
"glasgow",
"forthcoming",
"preparing",
"prospects",
"hoping",
"hopes",
"yukos",
"scots"
],
[
"investors",
"takeover",
"shareholder",
"shareholders",
"investor",
"deutsche",
"speculation",
"merger",
"securities",
"frankfurt"
]
] | 3.418174 | intfloat/e5-large-v2 | 0.975 | -0.317858 | 0.197561 | 0.902049 |
BBC News | S³_angular | 42 | 30 | [
[
"titles",
"title",
"opposed",
"publication",
"separate",
"principle",
"sequel",
"author",
"uncertain",
"planning"
],
[
"television",
"broadcast",
"viewers",
"dvds",
"broadcasting",
"shows",
"series",
"tv",
"channels",
"broadcaster"
],
[
"charity",
"compensation",
"aid",
"tsunami",
"globally",
"worldwide",
"global",
"world",
"funding",
"imf"
],
[
"xbox",
"consoles",
"games",
"playstation",
"sony",
"nintendo",
"console",
"psp",
"sequel",
"dvds"
],
[
"lansdowne",
"leigh",
"westminster",
"organisations",
"organisation",
"consortium",
"mps",
"milburn",
"clive",
"spokeswoman"
],
[
"interview",
"appeared",
"appears",
"happening",
"seen",
"subsequent",
"encounter",
"presence",
"visiting",
"interesting"
],
[
"scottish",
"scots",
"glasgow",
"scotland",
"edinburgh",
"margaret",
"celtic",
"campaigners",
"ryan",
"hunting"
],
[
"welsh",
"twickenham",
"wales",
"rugby",
"gareth",
"sullivan",
"ireland",
"wilkinson",
"jonny",
"williams"
],
[
"mortgage",
"economists",
"banks",
"banking",
"bank",
"forecasts",
"growth",
"economist",
"inflation",
"economy"
],
[
"norway",
"aren",
"netherlands",
"france",
"germany",
"italy",
"spain",
"countries",
"foreign",
"sweden"
],
[
"critics",
"disney",
"plays",
"directors",
"options",
"consumers",
"retailers",
"opinions",
"prince",
"files"
],
[
"films",
"cinema",
"screen",
"implications",
"training",
"director",
"screens",
"screening",
"film",
"movie"
],
[
"courts",
"lawyers",
"laws",
"court",
"judge",
"lords",
"law",
"detained",
"prisoner",
"appeals"
],
[
"interested",
"suggests",
"hoping",
"hopes",
"suggesting",
"interests",
"regarding",
"considering",
"concerns",
"wants"
],
[
"uefa",
"managers",
"goalkeeper",
"manager",
"arsenal",
"mourinho",
"striker",
"football",
"premiership",
"defender"
],
[
"disappointed",
"ignored",
"sadly",
"unacceptable",
"doubts",
"disappointing",
"disappointment",
"avoid",
"unfortunately",
"badly"
],
[
"closely",
"strikes",
"sunday",
"hospitals",
"coaches",
"repeated",
"strike",
"stores",
"women",
"tests"
],
[
"changing",
"changes",
"rebuild",
"landscape",
"revival",
"opening",
"smaller",
"slower",
"author",
"handle"
],
[
"yukos",
"gazprom",
"russian",
"russia",
"putin",
"khodorkovsky",
"oil",
"vladimir",
"mikhail",
"10bn"
],
[
"convinced",
"persuade",
"popularity",
"enthusiasm",
"lucrative",
"ambition",
"interviewed",
"interviews",
"influential",
"impressed"
],
[
"asking",
"yet",
"hardly",
"wanting",
"asked",
"she",
"herself",
"her",
"spokeswoman",
"denies"
],
[
"worried",
"losing",
"criticised",
"unfortunately",
"worrying",
"lose",
"disappointed",
"disappointing",
"sadly",
"failing"
],
[
"spammers",
"spam",
"mails",
"websites",
"privacy",
"mail",
"malicious",
"virus",
"viruses",
"security"
],
[
"actors",
"nominations",
"bafta",
"awards",
"oscars",
"nominees",
"nominee",
"nomination",
"nominated",
"winslet"
],
[
"auction",
"room",
"bills",
"resort",
"stream",
"suit",
"bill",
"session",
"monday",
"sir"
],
[
"frankfurt",
"deutsche",
"shareholders",
"european",
"europeans",
"german",
"negotiations",
"merger",
"europe",
"takeover"
],
[
"overseas",
"abroad",
"victims",
"movies",
"affairs",
"revenues",
"films",
"crime",
"hollywood",
"criminals"
],
[
"birmingham",
"agency",
"production",
"projects",
"wanted",
"agencies",
"newcastle",
"studios",
"u2",
"sky"
],
[
"factors",
"fi",
"zealand",
"regulators",
"agent",
"regulator",
"tries",
"academy",
"acts",
"spaniard"
],
[
"india",
"prospects",
"concerns",
"issues",
"investors",
"indian",
"competing",
"promising",
"aggressive",
"concerned"
]
] | 3.350451 | intfloat/e5-large-v2 | 0.966667 | -0.278097 | 0.224 | 0.918574 |
BBC News | S³_angular | 42 | 40 | [
[
"attacks",
"virus",
"threats",
"threat",
"infected",
"spammers",
"spam",
"security",
"viruses",
"secure"
],
[
"fears",
"shane",
"prix",
"publication",
"uncertain",
"prizes",
"prize",
"violence",
"hide",
"uncertainty"
],
[
"gb",
"finals",
"dvds",
"closed",
"retired",
"returns",
"closing",
"appeals",
"blockbuster",
"treasury"
],
[
"christmas",
"festive",
"hunt",
"alcohol",
"trophy",
"outdoor",
"hunting",
"restrictions",
"refuse",
"insisting"
],
[
"aid",
"poverty",
"20bn",
"aids",
"charity",
"imf",
"tsunami",
"africa",
"giving",
"worldwide"
],
[
"detained",
"prisoner",
"judges",
"lawyer",
"lawyers",
"courts",
"judge",
"lords",
"law",
"laws"
],
[
"technologies",
"future",
"announce",
"analysts",
"remarks",
"indicated",
"specifications",
"technology",
"broadcasters",
"commentators"
],
[
"earmarked",
"holmes",
"science",
"gear",
"string",
"rural",
"unveiled",
"2001",
"land",
"commons"
],
[
"photos",
"spokeswoman",
"spokesman",
"lots",
"newspapers",
"journalist",
"publicity",
"correspondent",
"pictures",
"spokesperson"
],
[
"manufacturers",
"brands",
"heineken",
"products",
"brand",
"replacements",
"competitors",
"woodward",
"replacing",
"product"
],
[
"association",
"reuters",
"zealand",
"months",
"links",
"sites",
"glasgow",
"november",
"friend",
"haven"
],
[
"nominees",
"nominated",
"awards",
"award",
"oscars",
"bafta",
"nominations",
"nomination",
"oscar",
"scorsese"
],
[
"scots",
"scottish",
"celtic",
"scotland",
"edinburgh",
"ryan",
"glasgow",
"ross",
"mac",
"stewart"
],
[
"lawsuit",
"allegations",
"controversy",
"fines",
"prosecutors",
"alleged",
"allegedly",
"doping",
"lawsuits",
"charges"
],
[
"gareth",
"wales",
"ireland",
"sullivan",
"welsh",
"rugby",
"wilkinson",
"twickenham",
"tackling",
"jonny"
],
[
"secretary",
"government",
"honours",
"contracts",
"contract",
"cabinet",
"mrs",
"35bn",
"premiership",
"5bn"
],
[
"mtv",
"critics",
"opinions",
"directors",
"retailers",
"dvd",
"viewers",
"consumers",
"options",
"plays"
],
[
"fa",
"quickly",
"sydney",
"sharply",
"melbourne",
"pointed",
"met",
"friends",
"500",
"barcelona"
],
[
"tv",
"viewers",
"broadcaster",
"ratings",
"television",
"comedian",
"mtv",
"broadcasters",
"comedy",
"broadcasting"
],
[
"rejected",
"abandoned",
"declined",
"spokeswoman",
"refused",
"hoped",
"denied",
"assistant",
"offered",
"ignored"
],
[
"bought",
"offered",
"dealing",
"deals",
"dealt",
"pledged",
"sold",
"deal",
"accepted",
"trusted"
],
[
"membership",
"withdrew",
"resigned",
"member",
"holder",
"constituency",
"candidate",
"elected",
"declined",
"ian"
],
[
"argentina",
"china",
"brazil",
"yuan",
"beijing",
"italy",
"chinese",
"countries",
"spain",
"spanish"
],
[
"old",
"older",
"aged",
"death",
"lifetime",
"lived",
"died",
"1987",
"1980s",
"1970s"
],
[
"apple",
"itunes",
"gadget",
"desktop",
"ipod",
"gadgets",
"mac",
"mini",
"computer",
"pocket"
],
[
"pensioners",
"replacements",
"meetings",
"candidates",
"strikes",
"unions",
"pensions",
"strike",
"pension",
"trials"
],
[
"u2",
"projects",
"social",
"publicity",
"squad",
"wanted",
"project",
"production",
"keane",
"runners"
],
[
"interviews",
"replied",
"afterwards",
"alongside",
"rumours",
"colleagues",
"colleague",
"blair",
"partners",
"peace"
],
[
"disappointing",
"worrying",
"criticised",
"criticism",
"bad",
"badly",
"disappointed",
"disappointment",
"unfortunately",
"sadly"
],
[
"announced",
"soon",
"producer",
"launching",
"announcement",
"launched",
"opening",
"forthcoming",
"releases",
"planned"
],
[
"pressures",
"released",
"portable",
"reduced",
"reducing",
"ran",
"sustain",
"returns",
"motivated",
"sustainable"
],
[
"microsoft",
"campaigning",
"internet",
"speech",
"headline",
"google",
"shareholders",
"amazon",
"website",
"web"
],
[
"campaign",
"coach",
"stations",
"teams",
"coaches",
"passengers",
"kids",
"newspapers",
"united",
"administration"
],
[
"enthusiasm",
"insisted",
"convinced",
"ambition",
"impressed",
"optimism",
"satisfied",
"delighted",
"confidence",
"persuade"
],
[
"regarding",
"believes",
"wants",
"concerns",
"suggesting",
"considering",
"hoping",
"suggests",
"concerned",
"interests"
],
[
"french",
"france",
"paris",
"nicolas",
"stable",
"recovered",
"parliamentary",
"window",
"spain",
"las"
],
[
"smith",
"represents",
"letwin",
"community",
"word",
"brand",
"picked",
"who",
"rap",
"values"
],
[
"unemployment",
"antonio",
"welfare",
"comedian",
"worker",
"anthony",
"stable",
"subsidies",
"local",
"vulnerable"
],
[
"american",
"americans",
"us",
"usa",
"america",
"dollar",
"dollars",
"washington",
"democrats",
"congress"
],
[
"camera",
"screens",
"screening",
"screen",
"putting",
"cameras",
"referee",
"considering",
"deputy",
"pointing"
]
] | 3.905976 | intfloat/e5-large-v2 | 0.9575 | -0.336627 | 0.205119 | 0.914469 |
BBC News | S³_angular | 42 | 50 | [
[
"suspected",
"receiving",
"received",
"imports",
"confirmed",
"card",
"agreement",
"secured",
"licence",
"payments"
],
[
"publication",
"iraq",
"jupiter",
"title",
"shane",
"author",
"publish",
"imf",
"titles",
"environment"
],
[
"decision",
"letter",
"dvd",
"decisions",
"disney",
"mtv",
"policies",
"retailers",
"vice",
"policy"
],
[
"spokesperson",
"spokesman",
"mother",
"presenter",
"alan",
"appearance",
"questioned",
"questions",
"pence",
"spokeswoman"
],
[
"20th",
"1987",
"1983",
"lived",
"death",
"died",
"old",
"century",
"1989",
"1991"
],
[
"gadgets",
"mini",
"itunes",
"desktop",
"ipod",
"apple",
"mac",
"laptops",
"computer",
"computers"
],
[
"originally",
"records",
"estimates",
"parts",
"initially",
"return",
"recall",
"download",
"version",
"repeated"
],
[
"chairman",
"delays",
"president",
"delayed",
"delay",
"waiting",
"wait",
"presidential",
"athens",
"election"
],
[
"irish",
"ireland",
"teams",
"immigration",
"team",
"immigrants",
"americans",
"hosts",
"chicago",
"overwhelming"
],
[
"messages",
"firm",
"message",
"firms",
"sales",
"principles",
"aside",
"album",
"principle",
"secretary"
],
[
"requests",
"waiting",
"wait",
"seeking",
"filing",
"status",
"account",
"claiming",
"request",
"citizens"
],
[
"dollars",
"dollar",
"americans",
"euros",
"us",
"usa",
"currency",
"american",
"america",
"euro"
],
[
"listeners",
"richard",
"owner",
"although",
"comments",
"careful",
"remarks",
"focusing",
"interests",
"woodward"
],
[
"staff",
"35bn",
"employees",
"departments",
"offices",
"facilities",
"angeles",
"journalists",
"accounts",
"sacked"
],
[
"citizens",
"pensioners",
"senior",
"elderly",
"insurance",
"taxes",
"age",
"tax",
"saving",
"goalkeeper"
],
[
"credit",
"securities",
"shareholders",
"financial",
"corporate",
"shareholder",
"sec",
"attorney",
"investor",
"investigation"
],
[
"welsh",
"france",
"rugby",
"cardiff",
"williams",
"wales",
"twickenham",
"midfield",
"sixth",
"letwin"
],
[
"soon",
"announcement",
"producer",
"release",
"launched",
"releases",
"planned",
"already",
"announced",
"announce"
],
[
"italian",
"argentina",
"parliament",
"mexico",
"spanish",
"western",
"sweden",
"reuters",
"las",
"texas"
],
[
"climate",
"handed",
"household",
"relay",
"deputy",
"gives",
"households",
"assistant",
"preparing",
"predicts"
],
[
"andre",
"kim",
"collins",
"martin",
"address",
"hasn",
"eric",
"kennedy",
"author",
"ellis"
],
[
"presenter",
"picked",
"smith",
"mike",
"pick",
"mikhail",
"representatives",
"don",
"represented",
"category"
],
[
"newspapers",
"wanted",
"renewed",
"favourites",
"campaigners",
"wants",
"ambitions",
"vera",
"larger",
"brian"
],
[
"disappointing",
"sadly",
"criticised",
"disappointment",
"disappointed",
"failed",
"avoid",
"failing",
"badly",
"cannot"
],
[
"matches",
"match",
"attend",
"each",
"organisers",
"labels",
"february",
"festival",
"showing",
"exhibition"
],
[
"loans",
"slowed",
"recession",
"bank",
"banking",
"banks",
"mortgage",
"housing",
"inflation",
"lending"
],
[
"beating",
"beaten",
"sunday",
"beat",
"howard",
"5bn",
"scoring",
"blues",
"hospitals",
"tougher"
],
[
"toshiba",
"dvds",
"dvd",
"consortium",
"alongside",
"technologies",
"formats",
"broadcasters",
"ibm",
"disc"
],
[
"insisted",
"achieved",
"certainly",
"believed",
"convinced",
"succeed",
"absolutely",
"achieve",
"successful",
"delighted"
],
[
"completed",
"breakthrough",
"seconds",
"longer",
"byrne",
"successive",
"scoring",
"finish",
"sideways",
"passed"
],
[
"arsenal",
"chelsea",
"brazilian",
"goalkeeper",
"mourinho",
"everton",
"barcelona",
"football",
"lisbon",
"jose"
],
[
"waiting",
"adopted",
"returning",
"italian",
"abroad",
"returned",
"landed",
"finally",
"spokeswoman",
"arrived"
],
[
"after",
"tense",
"afterwards",
"bomb",
"carefully",
"introduction",
"revolution",
"mood",
"careful",
"later"
],
[
"tax",
"howard",
"conservative",
"conservatives",
"tories",
"taxes",
"tory",
"liberal",
"campaigning",
"betting"
],
[
"doubles",
"tennis",
"tournament",
"roddick",
"wimbledon",
"federer",
"matches",
"court",
"courts",
"henman"
],
[
"sullivan",
"reduced",
"clarke",
"reducing",
"sustainable",
"sustain",
"joe",
"portable",
"driving",
"pcs"
],
[
"paris",
"olivier",
"prescott",
"korea",
"lisbon",
"edward",
"phillips",
"andre",
"amazon",
"thomas"
],
[
"early",
"ferdinand",
"aviator",
"national",
"tom",
"manchester",
"india",
"indian",
"dublin",
"agriculture"
],
[
"antonio",
"investment",
"reforms",
"argentina",
"reform",
"investments",
"backed",
"investors",
"invest",
"imf"
],
[
"imf",
"aids",
"africa",
"billion",
"gb",
"aid",
"globally",
"poverty",
"charity",
"worldwide"
],
[
"doping",
"athletics",
"athletes",
"olympics",
"drug",
"olympic",
"drugs",
"trials",
"prosecutors",
"fraud"
],
[
"award",
"announce",
"honours",
"awarded",
"recognised",
"correspondent",
"prize",
"announcement",
"recognition",
"announced"
],
[
"gazprom",
"mikhail",
"putin",
"russian",
"khodorkovsky",
"russia",
"yukos",
"oil",
"vladimir",
"norway"
],
[
"seen",
"looked",
"showing",
"showed",
"cameras",
"exhibition",
"attended",
"screening",
"watched",
"screens"
],
[
"controversy",
"feud",
"stance",
"aids",
"betting",
"rapper",
"angry",
"controversial",
"threatening",
"radical"
],
[
"betting",
"payment",
"marketing",
"bet",
"bidding",
"advertising",
"equipment",
"money",
"spent",
"receiving"
],
[
"german",
"germany",
"jobs",
"unemployment",
"worker",
"frankfurt",
"career",
"employment",
"job",
"economist"
],
[
"elvis",
"douglas",
"welfare",
"outcome",
"manchester",
"benefit",
"ferguson",
"ease",
"easier",
"stage"
],
[
"virus",
"infected",
"secure",
"viruses",
"security",
"spammers",
"protect",
"attacks",
"attack",
"breach"
],
[
"anthony",
"hop",
"choose",
"ranked",
"potter",
"falconer",
"choice",
"derby",
"ranks",
"venture"
]
] | 3.576301 | intfloat/e5-large-v2 | 0.934 | -0.334445 | 0.189505 | 0.905946 |
BBC News | S³_combined | 42 | 10 | [
[
"british",
"ofcom",
"uk",
"england",
"britons",
"britain",
"london",
"briton",
"twickenham",
"gb"
],
[
"stolen",
"knew",
"reported",
"complained",
"arrested",
"guilty",
"sued",
"spammers",
"caught",
"didn"
],
[
"bands",
"musical",
"directors",
"driscoll",
"advertising",
"films",
"songs",
"disney",
"band",
"critics"
],
[
"opposed",
"governments",
"restrictions",
"legislation",
"criticised",
"imposed",
"warnings",
"scrutiny",
"illegal",
"issues"
],
[
"regulators",
"agencies",
"registered",
"regulations",
"regulator",
"risks",
"regulation",
"agency",
"affected",
"bath"
],
[
"she",
"model",
"talks",
"her",
"agenda",
"commission",
"spokeswoman",
"herself",
"meetings",
"speech"
],
[
"increasing",
"recession",
"inflation",
"growth",
"increases",
"forecasts",
"economic",
"earnings",
"decline",
"economy"
],
[
"programs",
"warnings",
"youth",
"program",
"writer",
"writers",
"warning",
"dvds",
"careful",
"viewers"
],
[
"broadband",
"gadgets",
"technology",
"technologies",
"gadget",
"laptops",
"phones",
"handsets",
"mobiles",
"downloading"
],
[
"coaching",
"midfielder",
"mourinho",
"manager",
"managing",
"possibly",
"football",
"potentially",
"perhaps",
"goalkeeper"
]
] | 3.340383 | intfloat/e5-large-v2 | 0.99 | -0.325668 | 0.247194 | 0.917189 |
BBC News | S³_combined | 42 | 20 | [
[
"gone",
"resigned",
"mortgage",
"uk",
"westminster",
"abandoned",
"fewer",
"whilst",
"britons",
"sold"
],
[
"budgets",
"budget",
"costs",
"pensions",
"savings",
"saving",
"spending",
"cost",
"taxpayers",
"pay"
],
[
"lay",
"sharing",
"threw",
"divided",
"onto",
"relay",
"sued",
"adopt",
"adopted",
"channels"
],
[
"results",
"result",
"resulted",
"dominated",
"compared",
"passed",
"arrived",
"completed",
"achieved",
"scored"
],
[
"olympics",
"athletics",
"running",
"olympic",
"sprinter",
"medallist",
"athletes",
"competing",
"runners",
"hurdles"
],
[
"court",
"lawyers",
"constitutional",
"judge",
"courts",
"wait",
"parliament",
"detained",
"lawyer",
"consideration"
],
[
"ultimately",
"billion",
"bafta",
"wimbledon",
"championships",
"britain",
"concluded",
"british",
"birmingham",
"1970s"
],
[
"roddick",
"tennis",
"wimbledon",
"federer",
"court",
"courts",
"maria",
"matches",
"tournament",
"doubles"
],
[
"suggestions",
"expectations",
"analysts",
"interviews",
"contenders",
"itunes",
"networks",
"executives",
"shouldn",
"critics"
],
[
"abandoned",
"forgotten",
"scrapped",
"failure",
"closed",
"failed",
"unable",
"died",
"fails",
"slowdown"
],
[
"prosecution",
"attorney",
"prosecutors",
"lawsuits",
"lawsuit",
"alleged",
"charges",
"filing",
"accounting",
"allegations"
],
[
"managing",
"coaching",
"workforce",
"accounting",
"secretary",
"screening",
"intelligence",
"asset",
"linux",
"yen"
],
[
"sequel",
"nintendo",
"handheld",
"dvds",
"consoles",
"psp",
"sony",
"dvd",
"playstation",
"xbox"
],
[
"portable",
"mp3",
"downloaded",
"download",
"3g",
"mobile",
"midfield",
"electronics",
"band",
"technology"
],
[
"disappointment",
"criticised",
"badly",
"avoid",
"blamed",
"unfortunately",
"unacceptable",
"damaging",
"disappointed",
"disappointing"
],
[
"asia",
"holmes",
"country",
"legislation",
"countries",
"ireland",
"land",
"canada",
"japan",
"irish"
],
[
"blair",
"candidate",
"tories",
"election",
"campaigning",
"tory",
"conservative",
"campaign",
"candidates",
"conservatives"
],
[
"multimedia",
"bafta",
"mp3",
"ago",
"maria",
"mother",
"amazon",
"nokia",
"gazprom",
"public"
],
[
"forthcoming",
"scottish",
"glasgow",
"scotland",
"preparing",
"prospects",
"scots",
"hoping",
"yukos",
"returning"
],
[
"shareholders",
"merger",
"takeover",
"deutsche",
"shareholder",
"investor",
"investors",
"securities",
"frankfurt",
"proposals"
]
] | 3.426757 | intfloat/e5-large-v2 | 0.96 | -0.295028 | 0.216725 | 0.909317 |
BBC News | S³_combined | 42 | 30 | [
[
"title",
"author",
"sequel",
"titles",
"opposed",
"revised",
"publication",
"scorsese",
"asylum",
"creation"
],
[
"tv",
"dvd",
"television",
"dvds",
"viewers",
"broadcasting",
"broadcast",
"broadcaster",
"broadcasters",
"programmes"
],
[
"aid",
"tsunami",
"charity",
"funding",
"giving",
"imf",
"compensation",
"globally",
"helping",
"global"
],
[
"xbox",
"sony",
"playstation",
"nintendo",
"consoles",
"console",
"psp",
"dvd",
"games",
"dvds"
],
[
"organisations",
"consortium",
"leigh",
"organisation",
"westminster",
"lansdowne",
"mps",
"milburn",
"spokeswoman",
"everton"
],
[
"correspondent",
"recovered",
"encounter",
"appeared",
"interviewed",
"interview",
"subsequent",
"presence",
"incident",
"visiting"
],
[
"glasgow",
"scots",
"edinburgh",
"scotland",
"ryan",
"hunting",
"campaigners",
"celtic",
"scottish",
"margaret"
],
[
"ireland",
"twickenham",
"rugby",
"welsh",
"wales",
"gareth",
"williams",
"france",
"equipment",
"sullivan"
],
[
"economist",
"mortgage",
"bank",
"economists",
"banking",
"inflation",
"banks",
"economy",
"economics",
"growth"
],
[
"france",
"spain",
"italy",
"netherlands",
"norway",
"germany",
"japan",
"german",
"europe",
"french"
],
[
"critics",
"dvd",
"mtv",
"disney",
"directors",
"plays",
"consumers",
"files",
"dvds",
"file"
],
[
"film",
"screening",
"screen",
"films",
"cinema",
"director",
"movie",
"screens",
"training",
"referee"
],
[
"judge",
"court",
"detained",
"lawyers",
"laws",
"law",
"courts",
"prisoner",
"lawyer",
"constitutional"
],
[
"interests",
"wants",
"hoping",
"considering",
"regarding",
"interested",
"suggests",
"suggesting",
"tennis",
"concerns"
],
[
"goalkeeper",
"mourinho",
"football",
"uefa",
"chelsea",
"arsenal",
"manager",
"managers",
"striker",
"everton"
],
[
"disappointed",
"disappointing",
"sadly",
"ignored",
"avoid",
"unacceptable",
"unfortunately",
"rejected",
"badly",
"disappointment"
],
[
"together",
"hospitals",
"unions",
"closely",
"sunday",
"remember",
"coaches",
"repeated",
"women",
"strikes"
],
[
"smaller",
"revival",
"slower",
"author",
"landscape",
"rebuild",
"minister",
"norwich",
"handle",
"changing"
],
[
"gazprom",
"yukos",
"putin",
"khodorkovsky",
"russia",
"oil",
"russian",
"vladimir",
"mikhail",
"10bn"
],
[
"enthusiasm",
"satisfied",
"impressed",
"interviewed",
"convinced",
"lucrative",
"influential",
"popularity",
"achieved",
"persuade"
],
[
"yet",
"her",
"mrs",
"asking",
"actress",
"hardly",
"she",
"herself",
"denies",
"spokeswoman"
],
[
"disappointed",
"criticised",
"failure",
"worrying",
"lose",
"failing",
"disappointing",
"unfortunately",
"losing",
"sadly"
],
[
"mails",
"virus",
"spam",
"mail",
"spammers",
"viruses",
"privacy",
"infected",
"exploit",
"internet"
],
[
"nomination",
"nominees",
"oscars",
"actress",
"awards",
"award",
"nominee",
"bafta",
"nominations",
"nominated"
],
[
"bills",
"auction",
"cash",
"court",
"ofcom",
"fee",
"bank",
"stream",
"tv",
"room"
],
[
"german",
"europe",
"european",
"merger",
"negotiations",
"eu",
"deutsche",
"europeans",
"frankfurt",
"brussels"
],
[
"criminals",
"victims",
"revenues",
"films",
"abroad",
"hollywood",
"profits",
"jumped",
"movies",
"crime"
],
[
"birmingham",
"u2",
"bbc",
"manufacturing",
"agencies",
"keane",
"producing",
"production",
"projects",
"producer"
],
[
"regulator",
"regulators",
"zealand",
"fi",
"academy",
"spaniard",
"adopt",
"agent",
"czech",
"act"
],
[
"aggressive",
"competitive",
"issues",
"indian",
"competing",
"promising",
"encouraging",
"investors",
"india",
"concerns"
]
] | 3.36764 | intfloat/e5-large-v2 | 0.93 | -0.271974 | 0.235612 | 0.920512 |
BBC News | S³_combined | 42 | 40 | [
[
"infected",
"virus",
"spammers",
"viruses",
"spam",
"security",
"attacks",
"threats",
"secure",
"protect"
],
[
"publication",
"prix",
"prize",
"prizes",
"violence",
"author",
"shane",
"festival",
"anger",
"title"
],
[
"retired",
"gb",
"dvds",
"finals",
"treasury",
"closing",
"closed",
"dvd",
"debts",
"appeals"
],
[
"hunting",
"hunt",
"alcohol",
"outdoor",
"christmas",
"trophy",
"festive",
"fox",
"restrictions",
"refuse"
],
[
"aid",
"tsunami",
"charity",
"imf",
"poverty",
"africa",
"aids",
"giving",
"worldwide",
"poorest"
],
[
"detained",
"courts",
"prisoner",
"judges",
"judge",
"law",
"laws",
"lords",
"lawyer",
"lawyers"
],
[
"specifications",
"indicated",
"award",
"broadcasters",
"announce",
"nokia",
"technology",
"technologies",
"equipment",
"correspondent"
],
[
"hardware",
"rural",
"science",
"string",
"earmarked",
"holmes",
"gear",
"commons",
"eye",
"unveiled"
],
[
"spokesperson",
"publicity",
"phones",
"spokeswoman",
"lots",
"correspondent",
"spokesman",
"journalist",
"mobiles",
"photos"
],
[
"brands",
"manufacturers",
"heineken",
"products",
"brand",
"manufacturing",
"lions",
"companies",
"competitors",
"woodward"
],
[
"reuters",
"glasgow",
"zealand",
"cardiff",
"bristol",
"association",
"un",
"november",
"bath",
"contact"
],
[
"nominations",
"bafta",
"award",
"oscars",
"nominees",
"awards",
"awarded",
"nominated",
"oscar",
"nomination"
],
[
"glasgow",
"ryan",
"mac",
"celtic",
"campbell",
"scottish",
"edinburgh",
"scots",
"scotland",
"ross"
],
[
"lawsuits",
"fines",
"alleged",
"charges",
"allegations",
"doping",
"lawsuit",
"allegedly",
"offences",
"prosecutors"
],
[
"rugby",
"twickenham",
"gareth",
"welsh",
"ireland",
"wilkinson",
"wales",
"irish",
"injuries",
"sullivan"
],
[
"secretary",
"cabinet",
"pay",
"uk",
"fiscal",
"government",
"awarded",
"tories",
"mrs",
"eu"
],
[
"mtv",
"directors",
"dvds",
"dvd",
"critics",
"viewers",
"files",
"plays",
"letter",
"fund"
],
[
"milan",
"barcelona",
"met",
"fa",
"quickly",
"sydney",
"meetings",
"melbourne",
"lions",
"sharply"
],
[
"tv",
"broadcaster",
"television",
"viewers",
"broadcasting",
"ratings",
"comedian",
"broadcasters",
"comedy",
"mtv"
],
[
"declined",
"hoped",
"spokeswoman",
"rejected",
"refused",
"abandoned",
"wants",
"denied",
"offered",
"assistant"
],
[
"offered",
"dealt",
"deals",
"dealing",
"deal",
"bought",
"sold",
"accepted",
"purchase",
"agreement"
],
[
"membership",
"constituency",
"member",
"resigned",
"candidate",
"holder",
"withdrew",
"chairman",
"elected",
"ken"
],
[
"argentina",
"china",
"chinese",
"yuan",
"beijing",
"spanish",
"brazil",
"italy",
"barcelona",
"spain"
],
[
"old",
"older",
"aged",
"retired",
"death",
"died",
"lived",
"1980s",
"1970s",
"1987"
],
[
"itunes",
"ipod",
"gadgets",
"apple",
"mac",
"gadget",
"computer",
"desktop",
"mini",
"mp3"
],
[
"pensions",
"replacements",
"candidates",
"meeting",
"pensioners",
"meetings",
"strikes",
"unions",
"workers",
"pension"
],
[
"u2",
"projects",
"production",
"keane",
"project",
"publicity",
"squad",
"social",
"bbc",
"productions"
],
[
"blair",
"colleague",
"replied",
"interviews",
"encounter",
"alongside",
"interviewed",
"partners",
"colleagues",
"rumours"
],
[
"bad",
"slow",
"disappointed",
"disappointing",
"criticised",
"disappointment",
"slower",
"unfortunate",
"badly",
"unfortunately"
],
[
"launched",
"launching",
"announced",
"announcement",
"producer",
"animation",
"chancellor",
"entertainment",
"director",
"opening"
],
[
"sustain",
"reducing",
"youth",
"reduced",
"sustainable",
"portable",
"released",
"ran",
"returned",
"young"
],
[
"google",
"speech",
"microsoft",
"web",
"campaigning",
"internet",
"amazon",
"website",
"shareholders",
"campaign"
],
[
"united",
"kids",
"coaches",
"teams",
"administration",
"america",
"passengers",
"campaign",
"campaigning",
"operators"
],
[
"achieved",
"satisfied",
"impressed",
"convinced",
"optimism",
"enthusiasm",
"delighted",
"ambition",
"pleased",
"insisted"
],
[
"concerned",
"suggesting",
"hopefully",
"considering",
"believes",
"wants",
"suggests",
"hoping",
"concerns",
"regarding"
],
[
"nicolas",
"euro",
"euros",
"paris",
"france",
"spain",
"french",
"window",
"madrid",
"recovered"
],
[
"rap",
"represents",
"smith",
"word",
"community",
"letwin",
"rapper",
"kim",
"spokeswoman",
"who"
],
[
"worker",
"comedian",
"employed",
"unemployment",
"antonio",
"singer",
"working",
"welfare",
"richard",
"workers"
],
[
"dollars",
"us",
"washington",
"usa",
"americans",
"democrats",
"american",
"america",
"dollar",
"congress"
],
[
"camera",
"screening",
"screen",
"referee",
"deputy",
"cameras",
"screens",
"putting",
"pointing",
"cinema"
]
] | 4.062752 | intfloat/e5-large-v2 | 0.9375 | -0.316103 | 0.214108 | 0.908897 |
BBC News | S³_combined | 42 | 50 | [
[
"imports",
"card",
"franz",
"suspected",
"received",
"secured",
"agreement",
"licence",
"payments",
"netherlands"
],
[
"author",
"shane",
"title",
"iraq",
"jupiter",
"publication",
"imf",
"environmental",
"environment",
"publish"
],
[
"decision",
"mtv",
"disney",
"dvds",
"letter",
"dvd",
"aids",
"critics",
"tour",
"vice"
],
[
"spokeswoman",
"presenter",
"spokesperson",
"questioned",
"jewish",
"ken",
"mother",
"speaking",
"kim",
"pence"
],
[
"1987",
"1960s",
"1983",
"died",
"elvis",
"death",
"lived",
"old",
"dead",
"die"
],
[
"ipod",
"apple",
"itunes",
"gadgets",
"mac",
"desktop",
"mini",
"gadget",
"computer",
"laptops"
],
[
"recall",
"listen",
"originally",
"download",
"downloaded",
"records",
"estimates",
"initially",
"charts",
"parts"
],
[
"delayed",
"delays",
"chairman",
"president",
"presidential",
"wait",
"athens",
"delay",
"gazprom",
"election"
],
[
"americans",
"immigrants",
"teams",
"chicago",
"ireland",
"team",
"mobiles",
"immigration",
"hosts",
"irish"
],
[
"messages",
"message",
"firms",
"firm",
"album",
"secretary",
"sales",
"mails",
"principles",
"principle"
],
[
"file",
"waiting",
"citizens",
"site",
"account",
"requests",
"playing",
"filing",
"request",
"wait"
],
[
"us",
"euros",
"currency",
"americans",
"america",
"american",
"dollars",
"euro",
"dollar",
"usa"
],
[
"director",
"woodward",
"artist",
"focusing",
"actor",
"listeners",
"richard",
"listening",
"remarks",
"directors"
],
[
"staff",
"employees",
"departments",
"accounts",
"35bn",
"facilities",
"sacked",
"offices",
"fired",
"journalists"
],
[
"citizens",
"elderly",
"tax",
"pensioners",
"taxes",
"age",
"bankruptcy",
"insurance",
"fee",
"senior"
],
[
"investigation",
"accounting",
"sec",
"attorney",
"credit",
"securities",
"investor",
"shareholder",
"corporate",
"financial"
],
[
"rugby",
"wales",
"welsh",
"cardiff",
"france",
"williams",
"twickenham",
"wing",
"midfield",
"winger"
],
[
"announcement",
"announced",
"announce",
"producer",
"launched",
"launch",
"release",
"director",
"planned",
"launching"
],
[
"spanish",
"argentina",
"italian",
"mexico",
"madrid",
"italy",
"latin",
"texas",
"swiss",
"switzerland"
],
[
"deputy",
"climate",
"households",
"assistant",
"handed",
"campbell",
"relay",
"lending",
"household",
"predicts"
],
[
"african",
"kennedy",
"kim",
"ellis",
"collins",
"andre",
"martin",
"author",
"eric",
"dutch"
],
[
"mike",
"smith",
"presenter",
"mikhail",
"representatives",
"picked",
"category",
"pick",
"represented",
"culture"
],
[
"renewed",
"campaigners",
"ambitions",
"favourites",
"vera",
"nokia",
"brian",
"wants",
"larger",
"millennium"
],
[
"disappointing",
"criticised",
"disappointed",
"failed",
"disappointment",
"sadly",
"failing",
"badly",
"avoid",
"rejected"
],
[
"each",
"organisers",
"festival",
"attend",
"match",
"labels",
"matches",
"exhibition",
"meetings",
"meet"
],
[
"banking",
"loans",
"slowed",
"housing",
"mortgage",
"lending",
"inflation",
"recession",
"banks",
"bank"
],
[
"howard",
"beating",
"pace",
"scoring",
"beat",
"beaten",
"blues",
"windows",
"soul",
"injury"
],
[
"toshiba",
"consortium",
"dvd",
"ibm",
"dvds",
"broadcasters",
"disc",
"formats",
"technologies",
"alongside"
],
[
"satisfied",
"enthusiasm",
"convinced",
"achieve",
"achieved",
"successful",
"believed",
"delighted",
"insisted",
"succeed"
],
[
"passed",
"sideways",
"completed",
"unveiled",
"breakthrough",
"finished",
"byrne",
"seconds",
"successive",
"longer"
],
[
"barcelona",
"arsenal",
"mourinho",
"goalkeeper",
"chelsea",
"everton",
"lisbon",
"football",
"brazilian",
"jose"
],
[
"adopted",
"returning",
"returned",
"waiting",
"italian",
"landed",
"finally",
"abroad",
"arrived",
"spokeswoman"
],
[
"revolution",
"tense",
"sequel",
"bomb",
"square",
"introduction",
"afterwards",
"mood",
"comedy",
"terror"
],
[
"howard",
"tories",
"tax",
"conservative",
"tory",
"conservatives",
"taxes",
"campaigning",
"liberal",
"taxpayers"
],
[
"roddick",
"tournament",
"doubles",
"wimbledon",
"tennis",
"court",
"federer",
"courts",
"matches",
"henman"
],
[
"portable",
"sustainable",
"clarke",
"reduced",
"reducing",
"youth",
"iraq",
"pcs",
"driving",
"sustain"
],
[
"norwich",
"lisbon",
"edward",
"ellis",
"paris",
"amazon",
"olivier",
"phillips",
"andre",
"korea"
],
[
"dublin",
"indian",
"manchester",
"aviator",
"india",
"ferdinand",
"ireland",
"agriculture",
"tom",
"early"
],
[
"argentina",
"reform",
"reforms",
"investment",
"investments",
"antonio",
"imf",
"invest",
"spanish",
"restructuring"
],
[
"africa",
"aids",
"gb",
"imf",
"aid",
"charity",
"poverty",
"worldwide",
"billion",
"african"
],
[
"olympic",
"olympics",
"athletics",
"doping",
"athletes",
"drug",
"prosecutors",
"drugs",
"trial",
"trials"
],
[
"recognised",
"honours",
"award",
"awarded",
"announcement",
"announce",
"correspondent",
"oil",
"announced",
"prize"
],
[
"yukos",
"putin",
"russian",
"russia",
"khodorkovsky",
"gazprom",
"oil",
"vladimir",
"mikhail",
"norway"
],
[
"camera",
"showed",
"looked",
"seen",
"attended",
"watched",
"visitors",
"screening",
"exhibition",
"cameras"
],
[
"aids",
"violence",
"controversy",
"controversial",
"threatening",
"rapper",
"radical",
"betting",
"feud",
"angry"
],
[
"spend",
"advertising",
"marketing",
"funding",
"money",
"spent",
"equipment",
"spending",
"betting",
"payment"
],
[
"german",
"germany",
"unemployment",
"worker",
"employment",
"job",
"frankfurt",
"career",
"deutsche",
"jobs"
],
[
"ferguson",
"manchester",
"elvis",
"outcome",
"douglas",
"ease",
"welfare",
"hodgson",
"stage",
"franz"
],
[
"secure",
"infected",
"breach",
"spammers",
"viruses",
"virus",
"attacks",
"security",
"protect",
"attack"
],
[
"hop",
"derby",
"ranks",
"potter",
"venture",
"falconer",
"anthony",
"ranked",
"requirements",
"oliver"
]
] | 3.825573 | intfloat/e5-large-v2 | 0.92 | -0.328939 | 0.197894 | 0.921336 |
20 Newsgroups Preprocessed | BERTopic | 42 | 10 | [
[
"game",
"team",
"year",
"win",
"play",
"player",
"season",
"good",
"score",
"hit"
],
[
"time",
"good",
"people",
"file",
"work",
"make",
"system",
"find",
"problem",
"thing"
]
] | 45.145484 | intfloat/e5-large-v2 | 0.95 | 0.099572 | 0.238082 | 0.327736 |
20 Newsgroups Preprocessed | BERTopic | 42 | 20 | [
[
"game",
"team",
"year",
"win",
"play",
"player",
"season",
"good",
"score",
"hit"
],
[
"time",
"good",
"people",
"file",
"work",
"make",
"system",
"find",
"problem",
"thing"
]
] | 27.605216 | intfloat/e5-large-v2 | 0.95 | 0.099572 | 0.238082 | 0.332279 |
20 Newsgroups Preprocessed | BERTopic | 42 | 30 | [
[
"game",
"team",
"year",
"win",
"play",
"player",
"season",
"good",
"score",
"hit"
],
[
"time",
"good",
"people",
"file",
"work",
"make",
"system",
"find",
"problem",
"thing"
]
] | 27.73661 | intfloat/e5-large-v2 | 0.95 | 0.099572 | 0.238082 | 0.316777 |
20 Newsgroups Preprocessed | BERTopic | 42 | 40 | [
[
"game",
"team",
"year",
"win",
"play",
"player",
"season",
"good",
"score",
"hit"
],
[
"time",
"good",
"people",
"file",
"work",
"make",
"system",
"find",
"problem",
"thing"
]
] | 27.321661 | intfloat/e5-large-v2 | 0.95 | 0.099572 | 0.238082 | 0.321767 |
20 Newsgroups Preprocessed | BERTopic | 42 | 50 | [
[
"game",
"team",
"year",
"win",
"play",
"player",
"season",
"good",
"score",
"hit"
],
[
"time",
"good",
"people",
"file",
"work",
"make",
"system",
"find",
"problem",
"thing"
]
] | 27.318285 | intfloat/e5-large-v2 | 0.95 | 0.099572 | 0.238082 | 0.31484 |
20 Newsgroups Preprocessed | NMF | 42 | 10 | [
[
"version",
"format",
"color",
"bit",
"file",
"image",
"program",
"quality",
"display",
"software"
],
[
"thing",
"make",
"work",
"good",
"time",
"give",
"people",
"year",
"day",
"start"
],
[
"output",
"program",
"build",
"file",
"rule",
"section",
"entry",
"line",
"check",
"remark"
],
[
"server",
"widget",
"application",
"version",
"window",
"program",
"include",
"run",
"set",
"file"
],
[
"law",
"control",
"firearm",
"gun",
"directory",
"article",
"file",
"bill",
"weapon",
"crime"
],
[
"system",
"disk",
"support",
"drive",
"hard",
"card",
"controller",
"feature",
"bio",
"floppy"
],
[
"internet",
"system",
"information",
"privacy",
"user",
"computer",
"network",
"mail",
"email",
"key"
],
[
"earth",
"launch",
"year",
"planet",
"space",
"system",
"moon",
"satellite",
"solar",
"surface"
],
[
"box",
"power",
"connect",
"cable",
"ground",
"hot",
"circuit",
"wire",
"subject",
"require"
],
[
"send",
"image",
"graphic",
"package",
"mail",
"include",
"datum",
"object",
"server",
"format"
]
] | 2.654968 | intfloat/e5-large-v2 | 0.86 | 0.125153 | 0.149234 | 0.384849 |
20 Newsgroups Preprocessed | NMF | 42 | 20 | [
[
"format",
"version",
"bit",
"quality",
"image",
"color",
"file",
"program",
"display",
"free"
],
[
"leave",
"woman",
"kill",
"happen",
"time",
"armenian",
"live",
"start",
"people",
"child"
],
[
"output",
"check",
"file",
"line",
"open",
"entry",
"program",
"return",
"write",
"build"
],
[
"source",
"version",
"include",
"run",
"font",
"server",
"program",
"motif",
"subject",
"window"
],
[
"file",
"gun",
"firearm",
"control",
"law",
"bill",
"weapon",
"issue",
"crime",
"article"
],
[
"hard",
"disk",
"controller",
"drive",
"system",
"support",
"bio",
"card",
"feature",
"floppy"
],
[
"user",
"network",
"information",
"privacy",
"internet",
"email",
"computer",
"mail",
"address",
"service"
],
[
"system",
"orbit",
"moon",
"earth",
"solar",
"planet",
"surface",
"mission",
"sun",
"year"
],
[
"wire",
"box",
"hot",
"cable",
"ground",
"subject",
"circuit",
"connect",
"power",
"panel"
],
[
"server",
"mail",
"file",
"list",
"graphic",
"format",
"send",
"package",
"object",
"archive"
],
[
"system",
"file",
"disk",
"question",
"software",
"hard",
"problem",
"post",
"find",
"application"
],
[
"entry",
"file",
"program",
"build",
"section",
"rule",
"info",
"source",
"remark",
"number"
],
[
"report",
"patient",
"year",
"page",
"number",
"child",
"drug",
"health",
"disease",
"increase"
],
[
"team",
"game",
"year",
"player",
"play",
"win",
"season",
"draft",
"score",
"goal"
],
[
"set",
"widget",
"application",
"call",
"resource",
"window",
"type",
"default",
"display",
"datum"
],
[
"key",
"encryption",
"government",
"law",
"clipper",
"bit",
"device",
"chip",
"technology",
"enforcement"
],
[
"memory",
"window",
"card",
"support",
"software",
"server",
"high",
"system",
"color",
"version"
],
[
"commercial",
"space",
"market",
"satellite",
"year",
"launch",
"service",
"system",
"program",
"project"
],
[
"display",
"datum",
"package",
"software",
"tool",
"include",
"image",
"processing",
"analysis",
"user"
],
[
"make",
"work",
"people",
"president",
"question",
"good",
"give",
"thing",
"time",
"talk"
]
] | 3.609135 | intfloat/e5-large-v2 | 0.745 | 0.13058 | 0.151962 | 0.386364 |
20 Newsgroups Preprocessed | NMF | 42 | 30 | [
[
"image",
"file",
"color",
"version",
"program",
"bit",
"quality",
"format",
"display",
"free"
],
[
"woman",
"leave",
"start",
"kill",
"time",
"happen",
"child",
"neighbor",
"people",
"door"
],
[
"output",
"file",
"entry",
"check",
"line",
"program",
"open",
"return",
"build",
"write"
],
[
"widget",
"version",
"source",
"export",
"base",
"include",
"subject",
"motif",
"server",
"support"
],
[
"weapon",
"file",
"firearm",
"crime",
"gun",
"control",
"issue",
"bill",
"law",
"article"
],
[
"hard",
"disk",
"drive",
"controller",
"support",
"system",
"bio",
"feature",
"card",
"floppy"
],
[
"internet",
"information",
"privacy",
"email",
"user",
"mail",
"network",
"service",
"address",
"computer"
],
[
"moon",
"planet",
"earth",
"orbit",
"mission",
"system",
"solar",
"surface",
"sun",
"year"
],
[
"wire",
"ground",
"connect",
"box",
"cable",
"subject",
"circuit",
"hot",
"power",
"panel"
],
[
"file",
"send",
"graphic",
"mail",
"server",
"object",
"list",
"format",
"message",
"include"
],
[
"package",
"datum",
"model",
"call",
"format",
"system",
"software",
"graphic",
"display",
"program"
],
[
"entry",
"file",
"program",
"section",
"rule",
"build",
"info",
"source",
"remark",
"number"
],
[
"patient",
"number",
"child",
"page",
"report",
"drug",
"health",
"year",
"disease",
"increase"
],
[
"season",
"game",
"year",
"score",
"team",
"win",
"play",
"player",
"draft",
"goal"
],
[
"type",
"call",
"set",
"resource",
"application",
"widget",
"default",
"return",
"file",
"string"
],
[
"technology",
"encryption",
"key",
"law",
"government",
"clipper",
"chip",
"enforcement",
"device",
"public"
],
[
"server",
"card",
"system",
"support",
"high",
"software",
"version",
"memory",
"window",
"network"
],
[
"space",
"market",
"launch",
"satellite",
"commercial",
"year",
"service",
"technology",
"system",
"contract"
],
[
"datum",
"tool",
"user",
"image",
"include",
"processing",
"analysis",
"software",
"filter",
"line"
],
[
"russian",
"government",
"village",
"turkish",
"armenian",
"people",
"jewish",
"book",
"genocide",
"greek"
],
[
"bit",
"scsi",
"hardware",
"program",
"chip",
"build",
"machine",
"color",
"mode",
"bus"
],
[
"post",
"system",
"question",
"disk",
"file",
"problem",
"software",
"hard",
"newsgroup",
"find"
],
[
"decision",
"make",
"president",
"question",
"time",
"group",
"talk",
"general",
"work",
"option"
],
[
"people",
"work",
"job",
"make",
"good",
"thing",
"young",
"year",
"lot",
"school"
],
[
"government",
"program",
"administration",
"official",
"russian",
"fund",
"work",
"american",
"year",
"support"
],
[
"atheist",
"atheism",
"argument",
"religion",
"word",
"people",
"make",
"man",
"true",
"belief"
],
[
"attack",
"installation",
"secret",
"time",
"plan",
"military",
"war",
"ship",
"day",
"call"
],
[
"program",
"window",
"manager",
"server",
"font",
"run",
"subject",
"work",
"client",
"application"
],
[
"space",
"physical",
"theory",
"book",
"universe",
"star",
"time",
"light",
"speed",
"material"
],
[
"drive",
"driver",
"engine",
"good",
"tire",
"car",
"oil",
"time",
"brake",
"work"
]
] | 4.927839 | intfloat/e5-large-v2 | 0.703333 | 0.119203 | 0.152549 | 0.368512 |
20 Newsgroups Preprocessed | NMF | 42 | 40 | [
[
"format",
"file",
"version",
"image",
"color",
"quality",
"program",
"bit",
"display",
"free"
],
[
"happen",
"people",
"start",
"leave",
"woman",
"neighbor",
"kill",
"child",
"time",
"door"
],
[
"entry",
"file",
"program",
"output",
"line",
"open",
"return",
"build",
"check",
"write"
],
[
"version",
"source",
"include",
"export",
"motif",
"subject",
"base",
"widget",
"server",
"support"
],
[
"file",
"directory",
"bill",
"firearm",
"command",
"control",
"transfer",
"article",
"archive",
"report"
],
[
"drive",
"disk",
"card",
"support",
"hard",
"feature",
"bio",
"controller",
"system",
"floppy"
],
[
"information",
"email",
"privacy",
"user",
"internet",
"computer",
"network",
"mail",
"address",
"service"
],
[
"earth",
"moon",
"planet",
"solar",
"surface",
"system",
"orbit",
"mission",
"year",
"sun"
],
[
"ground",
"circuit",
"wire",
"subject",
"cable",
"box",
"hot",
"power",
"connect",
"panel"
],
[
"object",
"send",
"graphic",
"server",
"mail",
"format",
"list",
"message",
"include",
"archive"
],
[
"image",
"software",
"graphic",
"system",
"call",
"display",
"format",
"model",
"package",
"datum"
],
[
"rule",
"info",
"program",
"build",
"entry",
"file",
"section",
"source",
"remark",
"number"
],
[
"report",
"child",
"increase",
"number",
"year",
"age",
"rate",
"case",
"health",
"disease"
],
[
"draft",
"goal",
"year",
"score",
"season",
"win",
"game",
"player",
"play",
"team"
],
[
"application",
"widget",
"resource",
"call",
"default",
"set",
"type",
"return",
"datum",
"string"
],
[
"number",
"key",
"public",
"part",
"block",
"bit",
"encrypt",
"message",
"encryption",
"chip"
],
[
"system",
"high",
"support",
"software",
"card",
"server",
"version",
"memory",
"network",
"mouse"
],
[
"commercial",
"technology",
"system",
"satellite",
"year",
"service",
"space",
"market",
"launch",
"contract"
],
[
"tool",
"image",
"processing",
"user",
"include",
"analysis",
"datum",
"software",
"filter",
"line"
],
[
"government",
"history",
"turkish",
"greek",
"jewish",
"book",
"send",
"document",
"citizen",
"german"
],
[
"machine",
"hardware",
"build",
"chip",
"program",
"bit",
"color",
"scsi",
"mode",
"fast"
],
[
"question",
"software",
"disk",
"hard",
"problem",
"newsgroup",
"application",
"system",
"post",
"faq"
],
[
"people",
"work",
"job",
"thing",
"school",
"young",
"make",
"summer",
"country",
"lot"
],
[
"find",
"problem",
"scsi",
"driver",
"tape",
"work",
"time",
"window",
"run",
"error"
],
[
"official",
"administration",
"program",
"russian",
"government",
"fund",
"work",
"american",
"make",
"support"
],
[
"religion",
"atheist",
"belief",
"exist",
"atheism",
"people",
"religious",
"question",
"make",
"thing"
],
[
"military",
"ship",
"time",
"secret",
"war",
"plan",
"attack",
"installation",
"call",
"world"
],
[
"program",
"subject",
"manager",
"window",
"application",
"open",
"user",
"motif",
"run",
"font"
],
[
"give",
"thing",
"time",
"good",
"post",
"year",
"point",
"miss",
"make",
"excellent"
],
[
"car",
"brake",
"tire",
"oil",
"drive",
"engine",
"driver",
"wheel",
"speed",
"change"
],
[
"resource",
"server",
"file",
"client",
"font",
"program",
"start",
"run",
"find",
"display"
],
[
"light",
"space",
"book",
"time",
"physical",
"star",
"theory",
"universe",
"speed",
"material"
],
[
"person",
"dangerous",
"section",
"license",
"weapon",
"issue",
"carry",
"firearm",
"military",
"law"
],
[
"chip",
"protect",
"enforcement",
"government",
"encryption",
"technology",
"clipper",
"device",
"law",
"privacy"
],
[
"health",
"treatment",
"information",
"study",
"disease",
"page",
"patient",
"drug",
"medical",
"test"
],
[
"gun",
"law",
"crime",
"firearm",
"control",
"bill",
"weapon",
"study",
"rate",
"criminal"
],
[
"talk",
"make",
"question",
"work",
"general",
"president",
"decision",
"group",
"time",
"package"
],
[
"make",
"man",
"father",
"word",
"son",
"verse",
"state",
"god",
"day",
"speak"
],
[
"people",
"argument",
"homosexuality",
"homosexual",
"question",
"evidence",
"conclusion",
"true",
"word",
"law"
],
[
"russian",
"village",
"armenian",
"massacre",
"dead",
"muslim",
"genocide",
"soldier",
"people",
"population"
]
] | 5.385349 | intfloat/e5-large-v2 | 0.66 | 0.118685 | 0.151929 | 0.372113 |
20 Newsgroups Preprocessed | NMF | 42 | 50 | [
[
"program",
"quality",
"bit",
"version",
"format",
"free",
"color",
"display",
"file",
"image"
],
[
"leave",
"people",
"start",
"happen",
"time",
"woman",
"kill",
"neighbor",
"child",
"door"
],
[
"file",
"return",
"check",
"open",
"entry",
"output",
"program",
"line",
"build",
"write"
],
[
"server",
"motif",
"base",
"widget",
"support",
"include",
"version",
"export",
"source",
"subject"
],
[
"article",
"file",
"command",
"bill",
"control",
"archive",
"transfer",
"provide",
"directory",
"firearm"
],
[
"disk",
"drive",
"hard",
"controller",
"bio",
"support",
"feature",
"floppy",
"card",
"rom"
],
[
"network",
"user",
"email",
"electronic",
"computer",
"information",
"privacy",
"internet",
"mail",
"address"
],
[
"planet",
"surface",
"solar",
"moon",
"earth",
"sun",
"system",
"orbit",
"year",
"mile"
],
[
"wire",
"subject",
"circuit",
"ground",
"cable",
"box",
"connect",
"hot",
"power",
"panel"
],
[
"mail",
"post",
"list",
"send",
"request",
"mailing",
"day",
"information",
"article",
"motorcycle"
],
[
"filter",
"datum",
"tool",
"software",
"processing",
"analysis",
"user",
"image",
"include",
"line"
],
[
"rule",
"program",
"file",
"section",
"entry",
"build",
"source",
"info",
"remark",
"win"
],
[
"draft",
"player",
"year",
"season",
"game",
"team",
"play",
"division",
"hockey",
"win"
],
[
"make",
"president",
"question",
"decision",
"talk",
"general",
"time",
"package",
"option",
"yesterday"
],
[
"application",
"widget",
"call",
"set",
"resource",
"type",
"default",
"return",
"string",
"display"
],
[
"encrypt",
"public",
"bit",
"message",
"number",
"key",
"block",
"chip",
"part",
"attack"
],
[
"software",
"machine",
"system",
"version",
"user",
"application",
"list",
"type",
"base",
"port"
],
[
"contract",
"plan",
"space",
"commercial",
"year",
"market",
"project",
"service",
"satellite",
"launch"
],
[
"software",
"server",
"high",
"support",
"card",
"memory",
"network",
"mouse",
"version",
"graphic"
],
[
"thing",
"people",
"government",
"make",
"country",
"world",
"bear",
"free",
"live",
"lot"
],
[
"color",
"hardware",
"scsi",
"chip",
"bit",
"build",
"machine",
"program",
"mode",
"card"
],
[
"software",
"problem",
"post",
"faq",
"question",
"disk",
"hard",
"newsgroup",
"find",
"read"
],
[
"summer",
"program",
"work",
"young",
"school",
"year",
"job",
"time",
"community",
"learn"
],
[
"man",
"church",
"word",
"love",
"day",
"make",
"give",
"faith",
"find",
"speak"
],
[
"mail",
"include",
"format",
"object",
"server",
"package",
"file",
"image",
"graphic",
"send"
],
[
"atheism",
"belief",
"exist",
"religious",
"religion",
"question",
"atheist",
"existence",
"thing",
"make"
],
[
"plan",
"time",
"secret",
"war",
"ship",
"military",
"attack",
"call",
"installation",
"world"
],
[
"manager",
"application",
"window",
"program",
"subject",
"motif",
"user",
"screen",
"work",
"run"
],
[
"universe",
"theory",
"speed",
"physical",
"light",
"star",
"book",
"time",
"space",
"matter"
],
[
"car",
"brake",
"tire",
"oil",
"driver",
"speed",
"engine",
"wheel",
"change",
"make"
],
[
"find",
"client",
"server",
"program",
"resource",
"font",
"performance",
"start",
"load",
"file"
],
[
"fund",
"make",
"support",
"official",
"program",
"work",
"government",
"administration",
"russian",
"american"
],
[
"weapon",
"firearm",
"license",
"military",
"section",
"issue",
"dangerous",
"person",
"carry",
"division"
],
[
"encryption",
"chip",
"law",
"technology",
"protect",
"key",
"enforcement",
"device",
"government",
"clipper"
],
[
"health",
"disease",
"patient",
"number",
"year",
"report",
"child",
"page",
"increase",
"volume"
],
[
"problem",
"user",
"drug",
"patient",
"illegal",
"treatment",
"test",
"page",
"criminal",
"crime"
],
[
"site",
"service",
"posting",
"address",
"group",
"post",
"message",
"server",
"user",
"internet"
],
[
"son",
"verse",
"make",
"refer",
"god",
"father",
"state",
"man",
"earth",
"add"
],
[
"homosexual",
"sex",
"issue",
"law",
"christian",
"male",
"homosexuality",
"term",
"sexual",
"word"
],
[
"armenian",
"massacre",
"genocide",
"russian",
"population",
"muslim",
"village",
"people",
"soldier",
"dead"
],
[
"model",
"package",
"datum",
"graphic",
"display",
"call",
"software",
"format",
"set",
"program"
],
[
"law",
"crime",
"firearm",
"gun",
"control",
"rate",
"study",
"bill",
"weapon",
"criminal"
],
[
"send",
"government",
"jewish",
"greek",
"history",
"citizen",
"turkish",
"book",
"document",
"german"
],
[
"year",
"thing",
"good",
"excellent",
"miss",
"time",
"cover",
"lot",
"bad",
"win"
],
[
"police",
"israeli",
"jewish",
"report",
"information",
"arab",
"organization",
"group",
"anti",
"law"
],
[
"argument",
"false",
"occur",
"conclusion",
"evidence",
"true",
"form",
"make",
"valid",
"event"
],
[
"master",
"tape",
"drive",
"single",
"scsi",
"lead",
"open",
"ide",
"pin",
"mode"
],
[
"rocket",
"design",
"space",
"orbit",
"information",
"satellite",
"shuttle",
"mission",
"include",
"soviet"
],
[
"shot",
"period",
"win",
"game",
"goal",
"lead",
"play",
"power",
"score",
"blue"
],
[
"include",
"program",
"subject",
"open",
"run",
"version",
"font",
"display",
"change",
"start"
]
] | 9.10898 | intfloat/e5-large-v2 | 0.636 | 0.114135 | 0.149891 | 0.36584 |
20 Newsgroups Preprocessed | LDA | 42 | 10 | [
[
"year",
"people",
"drug",
"work",
"patient",
"study",
"problem",
"number",
"time",
"health"
],
[
"information",
"send",
"mail",
"key",
"list",
"chip",
"post",
"system",
"message",
"encryption"
],
[
"government",
"people",
"law",
"gun",
"make",
"state",
"crime",
"case",
"weapon",
"thing"
],
[
"buy",
"good",
"car",
"make",
"drive",
"work",
"sell",
"price",
"time",
"problem"
],
[
"win",
"team",
"good",
"game",
"player",
"year",
"play",
"season",
"make",
"run"
],
[
"armenian",
"people",
"kill",
"time",
"make",
"turkish",
"year",
"war",
"child",
"government"
],
[
"program",
"work",
"drive",
"card",
"version",
"problem",
"system",
"file",
"window",
"run"
],
[
"point",
"man",
"question",
"people",
"thing",
"word",
"make",
"time",
"good",
"read"
],
[
"time",
"game",
"goal",
"lead",
"back",
"play",
"period",
"night",
"day",
"point"
],
[
"launch",
"space",
"system",
"image",
"program",
"datum",
"mission",
"design",
"high",
"orbit"
]
] | 50.174856 | intfloat/e5-large-v2 | 0.71 | 0.069477 | 0.175032 | 0.255562 |
20 Newsgroups Preprocessed | LDA | 42 | 20 | [
[
"disease",
"study",
"drug",
"patient",
"health",
"doctor",
"medical",
"year",
"effect",
"problem"
],
[
"encryption",
"clipper",
"system",
"key",
"chip",
"phone",
"privacy",
"government",
"security",
"public"
],
[
"good",
"drive",
"make",
"ride",
"car",
"engine",
"buy",
"bike",
"time",
"mile"
],
[
"problem",
"make",
"power",
"work",
"good",
"wire",
"line",
"battery",
"ground",
"cable"
],
[
"team",
"hit",
"lose",
"year",
"make",
"time",
"run",
"good",
"player",
"win"
],
[
"call",
"time",
"year",
"day",
"leave",
"people",
"child",
"live",
"start",
"happen"
],
[
"monitor",
"problem",
"system",
"driver",
"card",
"disk",
"memory",
"video",
"mode",
"work"
],
[
"version",
"server",
"application",
"include",
"widget",
"support",
"resource",
"source",
"user",
"client"
],
[
"period",
"game",
"play",
"goal",
"score",
"team",
"fan",
"lead",
"shot",
"point"
],
[
"mission",
"earth",
"orbit",
"system",
"space",
"launch",
"satellite",
"cost",
"year",
"design"
],
[
"post",
"people",
"make",
"fire",
"thing",
"hear",
"kill",
"time",
"start",
"claim"
],
[
"year",
"job",
"president",
"make",
"work",
"good",
"team",
"time",
"people",
"talk"
],
[
"vote",
"gun",
"black",
"weapon",
"theory",
"law",
"bill",
"criminal",
"crime",
"firearm"
],
[
"program",
"file",
"window",
"problem",
"font",
"work",
"run",
"error",
"application",
"manager"
],
[
"jewish",
"turkish",
"war",
"people",
"israeli",
"armenian",
"government",
"kill",
"attack",
"village"
],
[
"word",
"thing",
"point",
"people",
"make",
"question",
"religion",
"good",
"church",
"man"
],
[
"color",
"bit",
"file",
"image",
"format",
"graphic",
"display",
"software",
"program",
"datum"
],
[
"court",
"freedom",
"government",
"state",
"make",
"case",
"law",
"people",
"power",
"free"
],
[
"mail",
"post",
"list",
"book",
"send",
"address",
"file",
"information",
"email",
"entry"
],
[
"scsi",
"offer",
"port",
"price",
"sell",
"drive",
"sale",
"modem",
"bus",
"good"
]
] | 91.287144 | intfloat/e5-large-v2 | 0.755 | 0.097752 | 0.168354 | 0.350257 |
20 Newsgroups Preprocessed | LDA | 42 | 30 | [
[
"drug",
"study",
"disease",
"patient",
"medical",
"treatment",
"year",
"doctor",
"number",
"effect"
],
[
"key",
"encryption",
"phone",
"government",
"chip",
"clipper",
"system",
"security",
"bit",
"privacy"
],
[
"engine",
"car",
"good",
"drive",
"price",
"buy",
"sell",
"make",
"front",
"mile"
],
[
"ground",
"cable",
"pin",
"battery",
"connect",
"wire",
"circuit",
"power",
"current",
"light"
],
[
"video",
"system",
"card",
"run",
"speed",
"monitor",
"bit",
"fast",
"color",
"mouse"
],
[
"money",
"make",
"pay",
"year",
"people",
"thing",
"insurance",
"private",
"spend",
"time"
],
[
"driver",
"price",
"work",
"problem",
"printer",
"memory",
"sale",
"software",
"disk",
"offer"
],
[
"server",
"program",
"application",
"file",
"window",
"run",
"font",
"display",
"set",
"problem"
],
[
"play",
"period",
"game",
"goal",
"blue",
"shot",
"lead",
"score",
"dog",
"red"
],
[
"cost",
"high",
"low",
"model",
"power",
"time",
"design",
"test",
"fuel",
"build"
],
[
"interested",
"hear",
"call",
"people",
"post",
"good",
"make",
"mail",
"company",
"order"
],
[
"question",
"president",
"decision",
"job",
"people",
"work",
"talk",
"make",
"time",
"press"
],
[
"theory",
"cover",
"black",
"science",
"book",
"white",
"material",
"physical",
"universe",
"vote"
],
[
"start",
"live",
"call",
"happen",
"day",
"home",
"back",
"leave",
"people",
"time"
],
[
"turkish",
"war",
"armenian",
"people",
"israeli",
"jewish",
"country",
"government",
"genocide",
"village"
],
[
"make",
"word",
"point",
"people",
"claim",
"question",
"argument",
"thing",
"belief",
"religion"
],
[
"work",
"control",
"input",
"datum",
"port",
"level",
"object",
"sound",
"create",
"application"
],
[
"people",
"law",
"government",
"state",
"case",
"make",
"arm",
"legal",
"court",
"power"
],
[
"read",
"find",
"problem",
"book",
"point",
"good",
"give",
"reference",
"text",
"make"
],
[
"hard",
"disk",
"system",
"drive",
"controller",
"modem",
"scsi",
"bus",
"tape",
"work"
],
[
"moon",
"mission",
"earth",
"satellite",
"orbit",
"year",
"system",
"launch",
"space",
"planet"
],
[
"time",
"work",
"ride",
"turn",
"bike",
"good",
"stop",
"motorcycle",
"friend",
"problem"
],
[
"address",
"list",
"system",
"message",
"post",
"information",
"mail",
"send",
"email",
"include"
],
[
"love",
"man",
"sin",
"people",
"life",
"make",
"good",
"die",
"give",
"day"
],
[
"water",
"people",
"time",
"church",
"group",
"member",
"post",
"year",
"feel",
"opinion"
],
[
"team",
"game",
"year",
"player",
"win",
"good",
"season",
"play",
"hit",
"make"
],
[
"kill",
"fire",
"people",
"child",
"gas",
"body",
"woman",
"dead",
"shoot",
"tank"
],
[
"police",
"gun",
"weapon",
"crime",
"firearm",
"criminal",
"law",
"batf",
"control",
"kill"
],
[
"entry",
"program",
"file",
"output",
"copy",
"build",
"section",
"line",
"rule",
"title"
],
[
"program",
"image",
"color",
"graphic",
"file",
"format",
"version",
"software",
"package",
"bit"
]
] | 115.376394 | intfloat/e5-large-v2 | 0.75 | 0.095733 | 0.166128 | 0.356293 |
20 Newsgroups Preprocessed | LDA | 42 | 40 | [
[
"medical",
"study",
"disease",
"patient",
"treatment",
"increase",
"health",
"effect",
"number",
"report"
],
[
"privacy",
"key",
"government",
"chip",
"clipper",
"security",
"system",
"encryption",
"technology",
"law"
],
[
"cover",
"good",
"make",
"light",
"clean",
"front",
"back",
"air",
"oil",
"drive"
],
[
"problem",
"tape",
"number",
"radio",
"phone",
"record",
"line",
"cop",
"find",
"call"
],
[
"run",
"fast",
"machine",
"speed",
"performance",
"slow",
"system",
"cpu",
"mhz",
"chip"
],
[
"private",
"insurance",
"time",
"spend",
"make",
"year",
"pay",
"money",
"care",
"ago"
],
[
"power",
"pin",
"battery",
"wire",
"cable",
"unit",
"ground",
"connect",
"circuit",
"switch"
],
[
"file",
"include",
"set",
"server",
"client",
"widget",
"resource",
"display",
"motif",
"application"
],
[
"play",
"period",
"game",
"goal",
"blue",
"shot",
"score",
"save",
"lead",
"penalty"
],
[
"system",
"mission",
"orbit",
"project",
"cost",
"launch",
"space",
"design",
"satellite",
"shuttle"
],
[
"lot",
"thing",
"call",
"hear",
"people",
"time",
"week",
"stuff",
"good",
"make"
],
[
"question",
"president",
"vote",
"decision",
"work",
"make",
"job",
"people",
"talk",
"time"
],
[
"material",
"book",
"theory",
"energy",
"science",
"universe",
"scientific",
"black",
"physical",
"explain"
],
[
"time",
"leave",
"people",
"start",
"home",
"happen",
"door",
"live",
"day",
"back"
],
[
"state",
"israeli",
"country",
"force",
"arab",
"war",
"jewish",
"village",
"russian",
"peace"
],
[
"belief",
"religion",
"argument",
"make",
"point",
"claim",
"people",
"question",
"evidence",
"exist"
],
[
"sound",
"level",
"tool",
"input",
"picture",
"datum",
"object",
"option",
"signal",
"work"
],
[
"state",
"law",
"case",
"make",
"power",
"legal",
"court",
"government",
"property",
"people"
],
[
"planet",
"sun",
"surface",
"water",
"moon",
"plane",
"temperature",
"earth",
"point",
"degree"
],
[
"ide",
"scsi",
"device",
"transfer",
"drive",
"bit",
"bus",
"modem",
"controller",
"datum"
],
[
"copy",
"launch",
"market",
"company",
"commercial",
"satellite",
"service",
"sell",
"price",
"current"
],
[
"offer",
"price",
"include",
"sale",
"sell",
"book",
"good",
"interested",
"manual",
"shipping"
],
[
"list",
"address",
"post",
"internet",
"send",
"email",
"message",
"information",
"mail",
"request"
],
[
"life",
"child",
"good",
"people",
"die",
"thing",
"make",
"death",
"human",
"man"
],
[
"control",
"time",
"people",
"study",
"year",
"bill",
"number",
"gun",
"make",
"rate"
],
[
"game",
"win",
"team",
"player",
"good",
"year",
"play",
"hit",
"season",
"make"
],
[
"child",
"burn",
"fire",
"compound",
"start",
"dog",
"tank",
"gas",
"claim",
"place"
],
[
"law",
"batf",
"criminal",
"crime",
"firearm",
"gun",
"police",
"weapon",
"government",
"citizen"
],
[
"entry",
"output",
"build",
"file",
"line",
"program",
"section",
"rule",
"write",
"info"
],
[
"image",
"file",
"format",
"software",
"bit",
"version",
"graphic",
"package",
"program",
"color"
],
[
"serial",
"message",
"number",
"random",
"bit",
"block",
"chip",
"key",
"part",
"encrypt"
],
[
"hard",
"software",
"drive",
"system",
"disk",
"problem",
"memory",
"port",
"work",
"card"
],
[
"problem",
"screen",
"font",
"run",
"window",
"program",
"work",
"manager",
"application",
"file"
],
[
"color",
"driver",
"apple",
"print",
"monitor",
"printer",
"card",
"video",
"mode",
"vga"
],
[
"post",
"group",
"drug",
"time",
"newsgroup",
"read",
"article",
"people",
"war",
"flame"
],
[
"bike",
"engine",
"car",
"ride",
"buy",
"time",
"make",
"drive",
"good",
"road"
],
[
"minority",
"greek",
"class",
"safety",
"live",
"issue",
"auto",
"insist",
"magazine",
"world"
],
[
"people",
"turkish",
"armenian",
"government",
"genocide",
"year",
"muslim",
"arm",
"history",
"page"
],
[
"word",
"man",
"day",
"sin",
"people",
"love",
"make",
"church",
"speak",
"faith"
],
[
"culture",
"woman",
"body",
"brain",
"child",
"claim",
"kill",
"jewish",
"man",
"blood"
]
] | 133.541383 | intfloat/e5-large-v2 | 0.7525 | 0.110197 | 0.164517 | 0.39826 |
20 Newsgroups Preprocessed | LDA | 42 | 50 | [
[
"patient",
"disease",
"study",
"doctor",
"medical",
"treatment",
"effect",
"food",
"health",
"research"
],
[
"encryption",
"government",
"system",
"chip",
"key",
"clipper",
"algorithm",
"bit",
"secure",
"security"
],
[
"good",
"condition",
"drive",
"power",
"air",
"excellent",
"cover",
"hard",
"thing",
"clean"
],
[
"buy",
"number",
"line",
"chip",
"make",
"battery",
"company",
"phone",
"cheap",
"sell"
],
[
"make",
"average",
"run",
"ball",
"hit",
"base",
"ability",
"double",
"performance",
"field"
],
[
"tax",
"pay",
"spend",
"insurance",
"private",
"year",
"make",
"cost",
"care",
"money"
],
[
"ground",
"pin",
"wire",
"box",
"power",
"connect",
"cable",
"unit",
"connector",
"remote"
],
[
"client",
"display",
"widget",
"server",
"set",
"motif",
"resource",
"function",
"include",
"application"
],
[
"play",
"period",
"blue",
"goal",
"lead",
"save",
"shot",
"sound",
"red",
"penalty"
],
[
"shuttle",
"design",
"fuel",
"flight",
"high",
"system",
"mission",
"stage",
"test",
"time"
],
[
"side",
"back",
"flame",
"great",
"guy",
"net",
"people",
"start",
"bank",
"couple"
],
[
"people",
"president",
"job",
"work",
"make",
"decision",
"talk",
"press",
"time",
"question"
],
[
"physical",
"matter",
"book",
"white",
"science",
"scientific",
"universe",
"material",
"theory",
"black"
],
[
"people",
"leave",
"home",
"happen",
"door",
"live",
"time",
"start",
"didn",
"woman"
],
[
"land",
"force",
"state",
"country",
"human",
"people",
"israeli",
"policy",
"military",
"citizen"
],
[
"make",
"argument",
"evidence",
"question",
"point",
"claim",
"exist",
"word",
"people",
"true"
],
[
"input",
"output",
"control",
"level",
"option",
"memory",
"signal",
"power",
"switch",
"current"
],
[
"case",
"freedom",
"legal",
"law",
"state",
"government",
"court",
"make",
"people",
"arm"
],
[
"dog",
"hot",
"cool",
"heat",
"oil",
"cold",
"point",
"temperature",
"face",
"fan"
],
[
"speed",
"bus",
"fast",
"scsi",
"drive",
"ide",
"device",
"mhz",
"bit",
"datum"
],
[
"satellite",
"technology",
"space",
"project",
"launch",
"program",
"development",
"cost",
"market",
"commercial"
],
[
"problem",
"time",
"work",
"good",
"student",
"check",
"find",
"turn",
"idea",
"stop"
],
[
"email",
"address",
"post",
"mail",
"list",
"information",
"newsgroup",
"message",
"internet",
"send"
],
[
"sin",
"man",
"child",
"love",
"life",
"die",
"make",
"son",
"good",
"death"
],
[
"long",
"year",
"month",
"time",
"water",
"ago",
"good",
"bad",
"past",
"week"
],
[
"game",
"win",
"play",
"good",
"team",
"year",
"player",
"season",
"lose",
"time"
],
[
"tank",
"claim",
"burn",
"gas",
"place",
"fire",
"eye",
"body",
"start",
"child"
],
[
"weapon",
"police",
"gun",
"criminal",
"crime",
"batf",
"people",
"law",
"firearm",
"death"
],
[
"entry",
"output",
"program",
"file",
"line",
"section",
"rule",
"copy",
"build",
"info"
],
[
"source",
"version",
"code",
"file",
"program",
"directory",
"software",
"package",
"find",
"include"
],
[
"hockey",
"sport",
"team",
"sign",
"fan",
"draft",
"ticket",
"bag",
"cover",
"man"
],
[
"hard",
"drive",
"software",
"modem",
"port",
"work",
"system",
"problem",
"disk",
"machine"
],
[
"window",
"problem",
"run",
"font",
"application",
"program",
"screen",
"manager",
"work",
"mouse"
],
[
"mode",
"monitor",
"apple",
"video",
"printer",
"driver",
"color",
"card",
"print",
"vga"
],
[
"drug",
"member",
"group",
"organization",
"secret",
"movement",
"war",
"church",
"time",
"intelligence"
],
[
"ride",
"engine",
"tire",
"make",
"car",
"drive",
"good",
"bike",
"buy",
"mile"
],
[
"auto",
"article",
"magazine",
"issue",
"world",
"real",
"safety",
"class",
"make",
"people"
],
[
"people",
"genocide",
"government",
"armenian",
"village",
"russian",
"turkish",
"muslim",
"greek",
"population"
],
[
"make",
"order",
"people",
"thing",
"call",
"church",
"day",
"time",
"speak",
"read"
],
[
"israeli",
"arab",
"kill",
"anti",
"woman",
"claim",
"attack",
"body",
"civilian",
"jewish"
],
[
"graphic",
"information",
"include",
"send",
"datum",
"software",
"computer",
"mail",
"package",
"system"
],
[
"man",
"sex",
"law",
"sexual",
"homosexuality",
"homosexual",
"gay",
"male",
"behavior",
"people"
],
[
"high",
"energy",
"model",
"low",
"figure",
"time",
"source",
"rate",
"light",
"power"
],
[
"child",
"weapon",
"law",
"firearm",
"question",
"bill",
"compound",
"cult",
"general",
"answer"
],
[
"display",
"format",
"color",
"program",
"image",
"quality",
"bit",
"convert",
"frame",
"size"
],
[
"planet",
"earth",
"surface",
"solar",
"moon",
"mission",
"orbit",
"sun",
"space",
"system"
],
[
"publish",
"bit",
"number",
"block",
"vote",
"difference",
"discussion",
"byte",
"propose",
"time"
],
[
"sale",
"offer",
"keyboard",
"price",
"sell",
"interested",
"include",
"shipping",
"ship",
"picture"
],
[
"book",
"write",
"reply",
"post",
"quote",
"read",
"article",
"send",
"find",
"word"
],
[
"belief",
"religious",
"people",
"atheism",
"atheist",
"religion",
"country",
"world",
"peace",
"make"
]
] | 152.27837 | intfloat/e5-large-v2 | 0.75 | 0.103376 | 0.169443 | 0.390529 |
20 Newsgroups Preprocessed | Top2Vec | 42 | 10 | [
[
"rule",
"discuss",
"sport",
"game",
"baseball",
"review",
"playoff",
"hockey",
"explain",
"league"
],
[
"additional",
"explain",
"computer",
"explanation",
"problem",
"review",
"follow",
"advice",
"instruction",
"reply"
]
] | 27.593331 | intfloat/e5-large-v2 | 0.9 | 0.013049 | 0.173361 | 0.266285 |
20 Newsgroups Preprocessed | Top2Vec | 42 | 20 | [
[
"rule",
"discuss",
"sport",
"game",
"baseball",
"review",
"playoff",
"hockey",
"explain",
"league"
],
[
"additional",
"explain",
"computer",
"explanation",
"problem",
"review",
"follow",
"advice",
"instruction",
"reply"
]
] | 27.63183 | intfloat/e5-large-v2 | 0.9 | 0.013049 | 0.173361 | 0.259362 |
20 Newsgroups Preprocessed | Top2Vec | 42 | 30 | [
[
"rule",
"discuss",
"sport",
"game",
"baseball",
"review",
"playoff",
"hockey",
"explain",
"league"
],
[
"additional",
"explain",
"computer",
"explanation",
"problem",
"review",
"follow",
"advice",
"instruction",
"reply"
]
] | 28.075856 | intfloat/e5-large-v2 | 0.9 | 0.013049 | 0.173361 | 0.261505 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.