Scheduled Commit

data/retrieval_battle-23cee96f-6768-4f6e-9a62-131dedf90618.jsonl

@@ -17,3 +17,9 @@
 {"tstamp": 1732228881.382, "task_type": "retrieval", "type": "leftvote", "models": ["", ""], "ip": "", "0_conv_id": "7703a6d7832f49698332041d621eb079", "0_model_name": "Salesforce/SFR-Embedding-2_R", "0_prompt": "July 15, 2016 coup attempt in Turkey", "0_output": [["July 15, 2016 coup attempt in Turkey", "2016 Turkish coup attempt\n\nOn 15 July 2016, a faction within the Turkish Armed Forces, organized as the Peace at Home Council, attempted a coup d'état against state institutions, including the government and president Recep Tayyip Erdoğan. They attempted to seize control of several places in Ankara, Istanbul, Marmaris and elsewhere, such as the Asian side entrance of the Bosphorus Bridge, but failed to do so after forces and civilians loyal to the state defeated them. The Council cited an erosion of secularism, elimination of democratic rule, disregard for human rights, and Turkey's loss of credibility in the international arena as reasons for the coup. The government said it had evidence the coup leaders were linked to the Gülen movement, which is designated as a terrorist organization by the Republic of Turkey and led by Fethullah Gülen, a Turkish businessman and a well-known Islamic scholar who lives in exile in Pennsylvania. The Turkish government alleged that Gülen was behind the coup (which Gülen denied) and that the United States was harboring him. Events surrounding the coup attempt and the purges in its aftermath reflect a complex power struggle between Islamist elites in Turkey."]], "0_corpus": "wikipedia", "1_conv_id": "06a4d40f174c4732b53d518473d1b1c1", "1_model_name": "BM25", "1_prompt": "July 15, 2016 coup attempt in Turkey", "1_output": [["July 15, 2016 coup attempt in Turkey", "Purges in Turkey following the 2016 Turkish coup attempt\n\nSince 2016, the government of Turkey has conducted a series of purges, enabled by a state of emergency in reaction to the failed coup attempt on 15 July that year. The purges began with the arrest of Turkish Armed Forces personnel reportedly linked to the coup attempt but arrests were expanded to include other elements of the Turkish military, as well as civil servants and private citizens. These later actions reflected a power struggle between secularist and Islamist political elites in Turkey, affected people who were not active in nor aware of the coup, but who the government claimed were connected with the Gülen movement, an opposition group which the government blamed for the coup. Possession of books authored by Gülen was considered valid evidence of such a connection and cause for arrest."]], "1_corpus": "wikipedia"}
 {"tstamp": 1732228903.8319, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "1c0b118068b7441a9481125700cc98eb", "0_model_name": "jinaai/jina-embeddings-v2-base-en", "0_prompt": "BDSM", "0_output": [["BDSM", "BDSM\n\nBDSM is a variety of often erotic practices or roleplaying involving bondage, discipline, dominance and submission, sadomasochism, and other related interpersonal dynamics. Given the wide range of practices, some of which may be engaged in by people who do not consider themselves to be practising BDSM, inclusion in the BDSM community or subculture often is said to depend on self-identification and shared experience.\nThe initialism BDSM is first recorded in a Usenet post from 1991, and is interpreted as a combination of the abbreviations B/D (Bondage and Discipline), D/s (Dominance and submission), and S/M (Sadism and Masochism). BDSM is now used as a catch-all phrase covering a wide range of activities, forms of interpersonal relationships, and distinct subcultures. BDSM communities generally welcome anyone with a non-normative streak who identifies with the community; this may include cross-dressers, body modification enthusiasts, animal roleplayers, rubber fetishists, and others."]], "0_corpus": "wikipedia", "1_conv_id": "b061ed39f3964ab894d1573eeb54dc0d", "1_model_name": "nomic-ai/nomic-embed-text-v1.5", "1_prompt": "BDSM", "1_output": [["BDSM", "BDSM\n\nBDSM is a variety of often erotic practices or roleplaying involving bondage, discipline, dominance and submission, sadomasochism, and other related interpersonal dynamics. Given the wide range of practices, some of which may be engaged in by people who do not consider themselves to be practising BDSM, inclusion in the BDSM community or subculture often is said to depend on self-identification and shared experience.\nThe initialism BDSM is first recorded in a Usenet post from 1991, and is interpreted as a combination of the abbreviations B/D (Bondage and Discipline), D/s (Dominance and submission), and S/M (Sadism and Masochism). BDSM is now used as a catch-all phrase covering a wide range of activities, forms of interpersonal relationships, and distinct subcultures. BDSM communities generally welcome anyone with a non-normative streak who identifies with the community; this may include cross-dressers, body modification enthusiasts, animal roleplayers, rubber fetishists, and others."]], "1_corpus": "wikipedia"}
 {"tstamp": 1732228966.0443, "task_type": "retrieval", "type": "leftvote", "models": ["", ""], "ip": "", "0_conv_id": "e625a7db097e48b08c14e38ea0e6de7f", "0_model_name": "intfloat/multilingual-e5-large-instruct", "0_prompt": "femdom and bdsm relationship", "0_output": [["femdom and bdsm relationship", "Outline of BDSM\n\nBDSM is a variety of erotic practices involving dominance and submission, roleplaying, restraint, and other interpersonal dynamics. Given the wide range of practices, some of which may be engaged in by people who do not consider themselves as practicing BDSM, inclusion in the BDSM community or subculture is usually dependent on self-identification and shared experience. Interest in BDSM can range from one-time experimentation to a lifestyle.\nThe following outline is provided as an overview of and topical guide to BDSM:\nEtymology\nThe term BDSM is a portmanteau of initialisms intended to encompass all of the following activities:\nBondage and discipline (B & D or B/D)\nDominance and submission (D & S or D/s) (including \"master and slave\" role-playing scenarios and ongoing relationship structures)\nSadism and masochism (S & M or S/M)\nBondage\nBy material\nJapanese bondage\nMetal bondage\nHandcuffs"]], "0_corpus": "wikipedia", "1_conv_id": "e76d12d7b57b46dc80c5e3255947f308", "1_model_name": "text-embedding-004", "1_prompt": "femdom and bdsm relationship", "1_output": [["femdom and bdsm relationship", "Female submission\n\nFemale submission or femsub is an activity or relationship in which a woman submits to the direction of a sexual partner or has her body used sexually by or for the sexual pleasure of her partner. The expression is often associated with BDSM, where submission to such activity is usually voluntarily and consensual. Submission usually involves a degree of trust by the woman in her partner. The dominant partner is usually a man, but can also be another woman, or there can be multiple dominant partners simultaneously. The submissive woman may derive sexual pleasure or emotional gratification from relinquishing (to varying degrees) control to (as well as satisfying) a trusted dominant partner.\nA 1985 study suggested that about 30% of participants in BDSM activities were females. A 2015 study indicates that 61.7% of females who are active in BDSM expressed a preference for a submissive role, 25.7% consider themselves a switch, while 12.6% prefer the dominant role but a more recent survey from 2017 indicates that women tend to self-identify as Submissive, Slave, Bottom, or Masochist (SSBM) and always perform submissive roles while men tend to self-identify as Dominant, Master, Top, or Sadist (DMTS) and always perform dominant roles."]], "1_corpus": "wikipedia"}
+{"tstamp": 1732229024.6113, "task_type": "retrieval", "type": "rightvote", "models": ["", ""], "ip": "", "0_conv_id": "e359cd9731394c659f458ec5d0e4f48a", "0_model_name": "sentence-transformers/all-MiniLM-L6-v2", "0_prompt": "Language families", "0_output": [["Language families", "Language family\n\nA language family is a group of languages related through descent from a common ancestral language or parental language, called the proto-language of that family. The term family is a metaphor borrowed from biology—with the tree model used in historical linguistics analogous to a family tree, or to phylogenetic trees of species used in evolutionary taxonomy. Linguists thus describe the daughter languages within a language family as being genetically related. The divergence of a proto-language into daughter languages typically occurs through geographical separation, with different regional dialects of the proto-language spoken by different speech communities undergoing different language changes and thus becoming distinct languages from each other."]], "0_corpus": "wikipedia", "1_conv_id": "efffb76fd47b4f7ab65b9b98b9cff625", "1_model_name": "text-embedding-004", "1_prompt": "Language families", "1_output": [["Language families", "Language family\n\nOne well-known example of a language family is the Romance languages, including Spanish, French, Italian, Portuguese, Romanian, and many others, all of which are descended from the Vulgar Latin spoken in the Roman Empire. The Romance family itself is part of the larger Indo-European family, which includes many other languages native to Europe and South Asia, such as English and Hindi, all believed to have descended from a common ancestor language spoken thousands of years ago known as Proto-Indo-European.\nThe language families with the most speakers are Indo-European and Sino-Tibetan—with the latter in large part due to the many speakers of Mandarin Chinese in China. A language family may contain any number of languages: some families, such as the Austronesian and Niger-Congo families, contain hundreds of different languages, while some languages, termed isolates, are not known to be related to any other languages and therefore constitute a family consisting of only one language.\nMembership of languages in a language family is established by research in comparative linguistics. Genealogically related languages can be identified by their shared retentions; that is, they share systematic similarities that cannot be explained as due to chance, or to effects of language contact (such as borrowing or convergence), and therefore must be features inherited from their shared common ancestor. However, some sets of languages may in fact be derived from a common ancestor but have diverged enough from each other that their relationship is no longer detectable; and some languages have not been studied in enough detail to be classified, and therefore their family membership is unknown."]], "1_corpus": "wikipedia"}
+{"tstamp": 1732229075.1102, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "41e5d47306764f03ae3e1e8114b9053c", "0_model_name": "intfloat/multilingual-e5-large-instruct", "0_prompt": "Julian assange", "0_output": [["Julian assange", "Julian Assange\n\nJulian Paul Assange ( ;  Hawkins; born 3 July 1971) is an Australian editor, publisher, and activist who founded WikiLeaks in 2006. He came to international attention in 2010 after WikiLeaks published a series of leaks from Chelsea Manning, a former United States Army intelligence analyst: footage of a U.S. airstrike in Baghdad, U.S. military logs from the Afghanistan and Iraq wars, and U.S. diplomatic cables. Assange has won multiple awards for publishing and journalism.\nAssange was raised in several towns in Australia until his family settled in Melbourne in his mid-teens. He became involved in the hacker community and was convicted for hacking in 1996. Following the establishment of WikiLeaks, Assange was its editor when it published the Bank Julius Baer documents, footage of the 2008 Tibetan unrest, and a report on political killings in Kenya with The Sunday Times."]], "0_corpus": "wikipedia", "1_conv_id": "8b5ea2b73e4f472ba3dde3404e7fee3a", "1_model_name": "Alibaba-NLP/gte-Qwen2-7B-instruct", "1_prompt": "Julian assange", "1_output": [["Julian assange", "Julian Assange\n\nJulian Paul Assange ( ;  Hawkins; born 3 July 1971) is an Australian editor, publisher, and activist who founded WikiLeaks in 2006. He came to international attention in 2010 after WikiLeaks published a series of leaks from Chelsea Manning, a former United States Army intelligence analyst: footage of a U.S. airstrike in Baghdad, U.S. military logs from the Afghanistan and Iraq wars, and U.S. diplomatic cables. Assange has won multiple awards for publishing and journalism.\nAssange was raised in several towns in Australia until his family settled in Melbourne in his mid-teens. He became involved in the hacker community and was convicted for hacking in 1996. Following the establishment of WikiLeaks, Assange was its editor when it published the Bank Julius Baer documents, footage of the 2008 Tibetan unrest, and a report on political killings in Kenya with The Sunday Times."]], "1_corpus": "wikipedia"}
+{"tstamp": 1732229166.9046, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "84ebf8c1a7534cc0a283100893fe8b09", "0_model_name": "voyage-multilingual-2", "0_prompt": "Baddeley's model of working memory", "0_output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "0_corpus": "wikipedia", "1_conv_id": "f4e9b3fc320649e382e0e437e8e4efae", "1_model_name": "embed-english-v3.0", "1_prompt": "Baddeley's model of working memory", "1_output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "1_corpus": "wikipedia"}
+{"tstamp": 1732229181.4518, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "20db37a161514b11b283c68bd80cc4dc", "0_model_name": "BM25", "0_prompt": "Baddeley's model of working memory", "0_output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "0_corpus": "wikipedia", "1_conv_id": "6231a029cd7e4e5d9230459b8b696129", "1_model_name": "voyage-multilingual-2", "1_prompt": "Baddeley's model of working memory", "1_output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "1_corpus": "wikipedia"}
+{"tstamp": 1732229213.5931, "task_type": "retrieval", "type": "rightvote", "models": ["", ""], "ip": "", "0_conv_id": "76a162d3df624890bfe6c2bf7eea9fc4", "0_model_name": "Salesforce/SFR-Embedding-2_R", "0_prompt": "working memory", "0_output": [["working memory", "Working memory\n\nWorking memory is a cognitive system with a limited capacity that can hold information temporarily. It is important for reasoning and the guidance of decision-making and behavior. Working memory is often used synonymously with short-term memory, but some theorists consider the two forms of memory distinct, assuming that working memory allows for the manipulation of stored information, whereas short-term memory only refers to the short-term storage of information. Working memory is a theoretical concept central to cognitive psychology, neuropsychology, and neuroscience."]], "0_corpus": "wikipedia", "1_conv_id": "c665709b31e74993a2f6ee6bc2bb9edc", "1_model_name": "BM25", "1_prompt": "working memory", "1_output": [["working memory", "Short-term memory\n\nSynaptic theory Various researchers have proposed that stimuli are coded in short-term memory using transmitter depletion. According to this hypothesis, a stimulus activates a spatial pattern of activity across neurons in a brain region. As these neurons fire, the available neurotransmitters are depleted. This depletion pattern represents stimulus information and functions as a memory trace. The memory trace decays over time as a consequence of neurotransmitter reuptake mechanisms that restore neurotransmitters to prior levels.\nRelationship with working memory\nThe relationship between short-term memory and working memory is described by various theories, but the two concepts are generally considered distinct. Neither holds information for long, but short-term memory is a simple store, while working memory allows it to be manipulated. Short-term memory is part of working memory, but is not the same thing.\nWorking memory refers to structures and processes used for temporarily storing and manipulating information. Working memory has been termed working attention. Working memory and attention together play a major role in the thought process. Short-term memory in general refers to the short-term storage of information, and it does not encompass memory manipulation or organization. Thus, while short-term memory components appear in working memory models, the concept of short-term memory is distinct from other concepts.\nWithin Baddeley's influential 1986 model of working memory two short-term storage mechanisms appear: the phonological loop and the visuospatial sketchpad. Most of the above research involves the phonological loop, because most of the work on short-term memory uses verbal material. Since the 1990s, however, research on visual short-term memory and spatial short-term memory has expanded."]], "1_corpus": "wikipedia"}
+{"tstamp": 1732229236.4019, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "4ea8c34aed2a42279cc3947c86c7dca6", "0_model_name": "nomic-ai/nomic-embed-text-v1.5", "0_prompt": "Which test was devised to determine whether robots can think?", "0_output": [["Which test was devised to determine whether robots can think?", "Turing test\n\nThe Turing test, originally called the imitation game by Alan Turing in 1950, is a test of a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human. Turing proposed that a human evaluator would judge natural language conversations between a human and a machine designed to generate human-like responses. The evaluator would be aware that one of the two partners in conversation was a machine, and all participants would be separated from one another. The conversation would be limited to a text-only channel, such as a computer keyboard and screen, so the result would not depend on the machine's ability to render words as speech. If the evaluator could not reliably tell the machine from the human, the machine would be said to have passed the test. The test results would not depend on the machine's ability to give correct answers to questions, only on how closely its answers resembled those a human would give. Since the Turing test is a test of indistinguishability in performance capacity, the verbal version generalizes naturally to all of human performance capacity, verbal as well as nonverbal (robotic)."]], "0_corpus": "wikipedia", "1_conv_id": "2d2b2eac434f46fe8d041a72ad07f30f", "1_model_name": "text-embedding-004", "1_prompt": "Which test was devised to determine whether robots can think?", "1_output": [["Which test was devised to determine whether robots can think?", "Turing test\n\nThe Turing test, originally called the imitation game by Alan Turing in 1950, is a test of a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human. Turing proposed that a human evaluator would judge natural language conversations between a human and a machine designed to generate human-like responses. The evaluator would be aware that one of the two partners in conversation was a machine, and all participants would be separated from one another. The conversation would be limited to a text-only channel, such as a computer keyboard and screen, so the result would not depend on the machine's ability to render words as speech. If the evaluator could not reliably tell the machine from the human, the machine would be said to have passed the test. The test results would not depend on the machine's ability to give correct answers to questions, only on how closely its answers resembled those a human would give. Since the Turing test is a test of indistinguishability in performance capacity, the verbal version generalizes naturally to all of human performance capacity, verbal as well as nonverbal (robotic)."]], "1_corpus": "wikipedia"}

data/retrieval_individual-23cee96f-6768-4f6e-9a62-131dedf90618.jsonl

@@ -58,3 +58,19 @@
 {"tstamp": 1732228894.6779, "task_type": "retrieval", "type": "chat", "model": "nomic-ai/nomic-embed-text-v1.5", "gen_params": {}, "start": 1732228894.413, "finish": 1732228894.6779, "ip": "", "conv_id": "b061ed39f3964ab894d1573eeb54dc0d", "model_name": "nomic-ai/nomic-embed-text-v1.5", "prompt": "BDSM", "output": [["BDSM", "BDSM\n\nBDSM is a variety of often erotic practices or roleplaying involving bondage, discipline, dominance and submission, sadomasochism, and other related interpersonal dynamics. Given the wide range of practices, some of which may be engaged in by people who do not consider themselves to be practising BDSM, inclusion in the BDSM community or subculture often is said to depend on self-identification and shared experience.\nThe initialism BDSM is first recorded in a Usenet post from 1991, and is interpreted as a combination of the abbreviations B/D (Bondage and Discipline), D/s (Dominance and submission), and S/M (Sadism and Masochism). BDSM is now used as a catch-all phrase covering a wide range of activities, forms of interpersonal relationships, and distinct subcultures. BDSM communities generally welcome anyone with a non-normative streak who identifies with the community; this may include cross-dressers, body modification enthusiasts, animal roleplayers, rubber fetishists, and others."]], "corpus": "wikipedia"}
 {"tstamp": 1732228947.5047, "task_type": "retrieval", "type": "chat", "model": "intfloat/multilingual-e5-large-instruct", "gen_params": {}, "start": 1732228947.1626, "finish": 1732228947.5047, "ip": "", "conv_id": "e625a7db097e48b08c14e38ea0e6de7f", "model_name": "intfloat/multilingual-e5-large-instruct", "prompt": "femdom and bdsm relationship", "output": [["femdom and bdsm relationship", "Outline of BDSM\n\nBDSM is a variety of erotic practices involving dominance and submission, roleplaying, restraint, and other interpersonal dynamics. Given the wide range of practices, some of which may be engaged in by people who do not consider themselves as practicing BDSM, inclusion in the BDSM community or subculture is usually dependent on self-identification and shared experience. Interest in BDSM can range from one-time experimentation to a lifestyle.\nThe following outline is provided as an overview of and topical guide to BDSM:\nEtymology\nThe term BDSM is a portmanteau of initialisms intended to encompass all of the following activities:\nBondage and discipline (B & D or B/D)\nDominance and submission (D & S or D/s) (including \"master and slave\" role-playing scenarios and ongoing relationship structures)\nSadism and masochism (S & M or S/M)\nBondage\nBy material\nJapanese bondage\nMetal bondage\nHandcuffs"]], "corpus": "wikipedia"}
 {"tstamp": 1732228947.5047, "task_type": "retrieval", "type": "chat", "model": "text-embedding-004", "gen_params": {}, "start": 1732228947.1626, "finish": 1732228947.5047, "ip": "", "conv_id": "e76d12d7b57b46dc80c5e3255947f308", "model_name": "text-embedding-004", "prompt": "femdom and bdsm relationship", "output": [["femdom and bdsm relationship", "Female submission\n\nFemale submission or femsub is an activity or relationship in which a woman submits to the direction of a sexual partner or has her body used sexually by or for the sexual pleasure of her partner. The expression is often associated with BDSM, where submission to such activity is usually voluntarily and consensual. Submission usually involves a degree of trust by the woman in her partner. The dominant partner is usually a man, but can also be another woman, or there can be multiple dominant partners simultaneously. The submissive woman may derive sexual pleasure or emotional gratification from relinquishing (to varying degrees) control to (as well as satisfying) a trusted dominant partner.\nA 1985 study suggested that about 30% of participants in BDSM activities were females. A 2015 study indicates that 61.7% of females who are active in BDSM expressed a preference for a submissive role, 25.7% consider themselves a switch, while 12.6% prefer the dominant role but a more recent survey from 2017 indicates that women tend to self-identify as Submissive, Slave, Bottom, or Masochist (SSBM) and always perform submissive roles while men tend to self-identify as Dominant, Master, Top, or Sadist (DMTS) and always perform dominant roles."]], "corpus": "wikipedia"}
+{"tstamp": 1732228999.1908, "task_type": "retrieval", "type": "chat", "model": "sentence-transformers/all-MiniLM-L6-v2", "gen_params": {}, "start": 1732228998.9123, "finish": 1732228999.1908, "ip": "", "conv_id": "e359cd9731394c659f458ec5d0e4f48a", "model_name": "sentence-transformers/all-MiniLM-L6-v2", "prompt": "Language families", "output": [["Language families", "Language family\n\nA language family is a group of languages related through descent from a common ancestral language or parental language, called the proto-language of that family. The term family is a metaphor borrowed from biology—with the tree model used in historical linguistics analogous to a family tree, or to phylogenetic trees of species used in evolutionary taxonomy. Linguists thus describe the daughter languages within a language family as being genetically related. The divergence of a proto-language into daughter languages typically occurs through geographical separation, with different regional dialects of the proto-language spoken by different speech communities undergoing different language changes and thus becoming distinct languages from each other."]], "corpus": "wikipedia"}
+{"tstamp": 1732228999.1908, "task_type": "retrieval", "type": "chat", "model": "text-embedding-004", "gen_params": {}, "start": 1732228998.9123, "finish": 1732228999.1908, "ip": "", "conv_id": "efffb76fd47b4f7ab65b9b98b9cff625", "model_name": "text-embedding-004", "prompt": "Language families", "output": [["Language families", "Language family\n\nOne well-known example of a language family is the Romance languages, including Spanish, French, Italian, Portuguese, Romanian, and many others, all of which are descended from the Vulgar Latin spoken in the Roman Empire. The Romance family itself is part of the larger Indo-European family, which includes many other languages native to Europe and South Asia, such as English and Hindi, all believed to have descended from a common ancestor language spoken thousands of years ago known as Proto-Indo-European.\nThe language families with the most speakers are Indo-European and Sino-Tibetan—with the latter in large part due to the many speakers of Mandarin Chinese in China. A language family may contain any number of languages: some families, such as the Austronesian and Niger-Congo families, contain hundreds of different languages, while some languages, termed isolates, are not known to be related to any other languages and therefore constitute a family consisting of only one language.\nMembership of languages in a language family is established by research in comparative linguistics. Genealogically related languages can be identified by their shared retentions; that is, they share systematic similarities that cannot be explained as due to chance, or to effects of language contact (such as borrowing or convergence), and therefore must be features inherited from their shared common ancestor. However, some sets of languages may in fact be derived from a common ancestor but have diverged enough from each other that their relationship is no longer detectable; and some languages have not been studied in enough detail to be classified, and therefore their family membership is unknown."]], "corpus": "wikipedia"}
+{"tstamp": 1732229050.0635, "task_type": "retrieval", "type": "chat", "model": "intfloat/multilingual-e5-large-instruct", "gen_params": {}, "start": 1732229047.8096, "finish": 1732229050.0635, "ip": "", "conv_id": "41e5d47306764f03ae3e1e8114b9053c", "model_name": "intfloat/multilingual-e5-large-instruct", "prompt": "Julian assange", "output": [["Julian assange", "Julian Assange\n\nJulian Paul Assange ( ;  Hawkins; born 3 July 1971) is an Australian editor, publisher, and activist who founded WikiLeaks in 2006. He came to international attention in 2010 after WikiLeaks published a series of leaks from Chelsea Manning, a former United States Army intelligence analyst: footage of a U.S. airstrike in Baghdad, U.S. military logs from the Afghanistan and Iraq wars, and U.S. diplomatic cables. Assange has won multiple awards for publishing and journalism.\nAssange was raised in several towns in Australia until his family settled in Melbourne in his mid-teens. He became involved in the hacker community and was convicted for hacking in 1996. Following the establishment of WikiLeaks, Assange was its editor when it published the Bank Julius Baer documents, footage of the 2008 Tibetan unrest, and a report on political killings in Kenya with The Sunday Times."]], "corpus": "wikipedia"}
+{"tstamp": 1732229050.0635, "task_type": "retrieval", "type": "chat", "model": "Alibaba-NLP/gte-Qwen2-7B-instruct", "gen_params": {}, "start": 1732229047.8096, "finish": 1732229050.0635, "ip": "", "conv_id": "8b5ea2b73e4f472ba3dde3404e7fee3a", "model_name": "Alibaba-NLP/gte-Qwen2-7B-instruct", "prompt": "Julian assange", "output": [["Julian assange", "Julian Assange\n\nJulian Paul Assange ( ;  Hawkins; born 3 July 1971) is an Australian editor, publisher, and activist who founded WikiLeaks in 2006. He came to international attention in 2010 after WikiLeaks published a series of leaks from Chelsea Manning, a former United States Army intelligence analyst: footage of a U.S. airstrike in Baghdad, U.S. military logs from the Afghanistan and Iraq wars, and U.S. diplomatic cables. Assange has won multiple awards for publishing and journalism.\nAssange was raised in several towns in Australia until his family settled in Melbourne in his mid-teens. He became involved in the hacker community and was convicted for hacking in 1996. Following the establishment of WikiLeaks, Assange was its editor when it published the Bank Julius Baer documents, footage of the 2008 Tibetan unrest, and a report on political killings in Kenya with The Sunday Times."]], "corpus": "wikipedia"}
+{"tstamp": 1732229132.2075, "task_type": "retrieval", "type": "chat", "model": "GritLM/GritLM-7B", "gen_params": {}, "start": 1732229129.9676, "finish": 1732229132.2075, "ip": "", "conv_id": "371c7cd3d30e40239cfa8ca800dd0cf5", "model_name": "GritLM/GritLM-7B", "prompt": "executive memory", "output": [["executive memory", "Baddeley's model of working memory\n\nCentral executive The central executive is a flexible system responsible for the control and regulation of cognitive processes. It directs focus and targets information, making working memory and long-term memory work together. It can be thought of as a supervisory system that controls cognitive processes, making sure the short-term store is actively working, and intervenes when they go astray and prevents distractions.\nIt has the following functions:\nUpdating and coding incoming information and replacing old information\nBinding information from a number of sources into coherent episodes\nCoordination of the slave systems\nShifting between tasks or retrieval strategies\nInhibition, suppressing dominant or automatic responses\nSelective attention\nThe central executive has two main systems: the visuo-spatial sketchpad, for visual information, and the phonological loop, for verbal information.\nUsing the dual-task paradigm, Baddeley and Erses have found, for instance, that patients with Alzheimer's dementia are impaired when performing multiple tasks simultaneously, even when the difficulty of the individual tasks is adapted to their abilities.   Two tasks include a memory tasks and a tracking task. Individual actions are completed well, but as the Alzheimer's becomes more prominent in a patient, performing two or more actions becomes more and more difficult.  This research has shown the deteriorating of the central executive in individuals with Alzheimer's."]], "corpus": "wikipedia"}
+{"tstamp": 1732229132.2075, "task_type": "retrieval", "type": "chat", "model": "embed-english-v3.0", "gen_params": {}, "start": 1732229129.9676, "finish": 1732229132.2075, "ip": "", "conv_id": "ae643e98ff3e4389b5f11c07b7c7d989", "model_name": "embed-english-v3.0", "prompt": "executive memory", "output": [["executive memory", "Neuropsychological test\n\nMemory Memory is a very broad function which includes several distinct abilities, all of which can be selectively impaired and require individual testing. There is disagreement as to the number of memory systems, depending on the psychological perspective taken. From a clinical perspective, a view of five distinct types of memory, is in most cases sufficient. Semantic memory and episodic memory (collectively called declarative memory or explicit memory); procedural memory and priming or perceptual learning (collectively called non-declarative memory or implicit memory) all four of which are long term memory systems; and working memory or short term memory. Semantic memory is memory for facts, episodic memory is autobiographical memory, procedural memory is memory for the performance of skills, priming is memory facilitated by prior exposure to a stimulus and working memory is a form of short term memory for information manipulation.}\nBenton Visual Retention Test\nCalifornia Verbal Learning Test\nCambridge Prospective Memory Test (CAMPROMPT)\nGollin figure test\nMemory Assessment Scales (MAS)\nRey Auditory Verbal Learning Test\nRivermead Behavioural Memory Test\nTest of Memory and Learning (TOMAL)\nMental Attributes Profiling System\nWechsler Memory Scale (WMS)\nLanguage\nLanguage functions include speech, reading and writing, all of which can be selectively impaired.}\nBoston Diagnostic Aphasia Examination\nBoston Naming Test\nComprehensive Aphasia Test (CAT)\nMultilingual Aphasia Examination\nExecutive function\nExecutive functions is an umbrella term for a various cognitive processes and sub-processes. The executive functions include: problem solving, planning, organizational skills, selective attention, inhibitory control and some aspects of short term memory.}"]], "corpus": "wikipedia"}
+{"tstamp": 1732229162.6431, "task_type": "retrieval", "type": "chat", "model": "voyage-multilingual-2", "gen_params": {}, "start": 1732229162.0962, "finish": 1732229162.6431, "ip": "", "conv_id": "84ebf8c1a7534cc0a283100893fe8b09", "model_name": "voyage-multilingual-2", "prompt": "Baddeley's model of working memory", "output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "corpus": "wikipedia"}
+{"tstamp": 1732229162.6431, "task_type": "retrieval", "type": "chat", "model": "embed-english-v3.0", "gen_params": {}, "start": 1732229162.0962, "finish": 1732229162.6431, "ip": "", "conv_id": "f4e9b3fc320649e382e0e437e8e4efae", "model_name": "embed-english-v3.0", "prompt": "Baddeley's model of working memory", "output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "corpus": "wikipedia"}
+{"tstamp": 1732229179.0471, "task_type": "retrieval", "type": "chat", "model": "BM25", "gen_params": {}, "start": 1732229176.1247, "finish": 1732229179.0471, "ip": "", "conv_id": "20db37a161514b11b283c68bd80cc4dc", "model_name": "BM25", "prompt": "Baddeley's model of working memory", "output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "corpus": "wikipedia"}
+{"tstamp": 1732229179.0471, "task_type": "retrieval", "type": "chat", "model": "voyage-multilingual-2", "gen_params": {}, "start": 1732229176.1247, "finish": 1732229179.0471, "ip": "", "conv_id": "6231a029cd7e4e5d9230459b8b696129", "model_name": "voyage-multilingual-2", "prompt": "Baddeley's model of working memory", "output": [["Baddeley's model of working memory", "Baddeley's model of working memory\n\nBaddeley's model of working memory is a model of human memory proposed by Alan Baddeley and Graham Hitch in 1974, in an attempt to present a more accurate model of primary memory (often referred to as short-term memory). Working memory splits primary memory into multiple components, rather than considering it to be a single, unified construct."]], "corpus": "wikipedia"}
+{"tstamp": 1732229196.6792, "task_type": "retrieval", "type": "chat", "model": "Salesforce/SFR-Embedding-2_R", "gen_params": {}, "start": 1732229196.2325, "finish": 1732229196.6792, "ip": "", "conv_id": "76a162d3df624890bfe6c2bf7eea9fc4", "model_name": "Salesforce/SFR-Embedding-2_R", "prompt": "working memory", "output": [["working memory", "Working memory\n\nWorking memory is a cognitive system with a limited capacity that can hold information temporarily. It is important for reasoning and the guidance of decision-making and behavior. Working memory is often used synonymously with short-term memory, but some theorists consider the two forms of memory distinct, assuming that working memory allows for the manipulation of stored information, whereas short-term memory only refers to the short-term storage of information. Working memory is a theoretical concept central to cognitive psychology, neuropsychology, and neuroscience."]], "corpus": "wikipedia"}
+{"tstamp": 1732229196.6792, "task_type": "retrieval", "type": "chat", "model": "BM25", "gen_params": {}, "start": 1732229196.2325, "finish": 1732229196.6792, "ip": "", "conv_id": "c665709b31e74993a2f6ee6bc2bb9edc", "model_name": "BM25", "prompt": "working memory", "output": [["working memory", "Short-term memory\n\nSynaptic theory Various researchers have proposed that stimuli are coded in short-term memory using transmitter depletion. According to this hypothesis, a stimulus activates a spatial pattern of activity across neurons in a brain region. As these neurons fire, the available neurotransmitters are depleted. This depletion pattern represents stimulus information and functions as a memory trace. The memory trace decays over time as a consequence of neurotransmitter reuptake mechanisms that restore neurotransmitters to prior levels.\nRelationship with working memory\nThe relationship between short-term memory and working memory is described by various theories, but the two concepts are generally considered distinct. Neither holds information for long, but short-term memory is a simple store, while working memory allows it to be manipulated. Short-term memory is part of working memory, but is not the same thing.\nWorking memory refers to structures and processes used for temporarily storing and manipulating information. Working memory has been termed working attention. Working memory and attention together play a major role in the thought process. Short-term memory in general refers to the short-term storage of information, and it does not encompass memory manipulation or organization. Thus, while short-term memory components appear in working memory models, the concept of short-term memory is distinct from other concepts.\nWithin Baddeley's influential 1986 model of working memory two short-term storage mechanisms appear: the phonological loop and the visuospatial sketchpad. Most of the above research involves the phonological loop, because most of the work on short-term memory uses verbal material. Since the 1990s, however, research on visual short-term memory and spatial short-term memory has expanded."]], "corpus": "wikipedia"}
+{"tstamp": 1732229231.2282, "task_type": "retrieval", "type": "chat", "model": "nomic-ai/nomic-embed-text-v1.5", "gen_params": {}, "start": 1732229230.9398, "finish": 1732229231.2282, "ip": "", "conv_id": "4ea8c34aed2a42279cc3947c86c7dca6", "model_name": "nomic-ai/nomic-embed-text-v1.5", "prompt": "Which test was devised to determine whether robots can think?", "output": [["Which test was devised to determine whether robots can think?", "Turing test\n\nThe Turing test, originally called the imitation game by Alan Turing in 1950, is a test of a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human. Turing proposed that a human evaluator would judge natural language conversations between a human and a machine designed to generate human-like responses. The evaluator would be aware that one of the two partners in conversation was a machine, and all participants would be separated from one another. The conversation would be limited to a text-only channel, such as a computer keyboard and screen, so the result would not depend on the machine's ability to render words as speech. If the evaluator could not reliably tell the machine from the human, the machine would be said to have passed the test. The test results would not depend on the machine's ability to give correct answers to questions, only on how closely its answers resembled those a human would give. Since the Turing test is a test of indistinguishability in performance capacity, the verbal version generalizes naturally to all of human performance capacity, verbal as well as nonverbal (robotic)."]], "corpus": "wikipedia"}
+{"tstamp": 1732229231.2282, "task_type": "retrieval", "type": "chat", "model": "text-embedding-004", "gen_params": {}, "start": 1732229230.9398, "finish": 1732229231.2282, "ip": "", "conv_id": "2d2b2eac434f46fe8d041a72ad07f30f", "model_name": "text-embedding-004", "prompt": "Which test was devised to determine whether robots can think?", "output": [["Which test was devised to determine whether robots can think?", "Turing test\n\nThe Turing test, originally called the imitation game by Alan Turing in 1950, is a test of a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human. Turing proposed that a human evaluator would judge natural language conversations between a human and a machine designed to generate human-like responses. The evaluator would be aware that one of the two partners in conversation was a machine, and all participants would be separated from one another. The conversation would be limited to a text-only channel, such as a computer keyboard and screen, so the result would not depend on the machine's ability to render words as speech. If the evaluator could not reliably tell the machine from the human, the machine would be said to have passed the test. The test results would not depend on the machine's ability to give correct answers to questions, only on how closely its answers resembled those a human would give. Since the Turing test is a test of indistinguishability in performance capacity, the verbal version generalizes naturally to all of human performance capacity, verbal as well as nonverbal (robotic)."]], "corpus": "wikipedia"}
+{"tstamp": 1732229264.1403, "task_type": "retrieval", "type": "chat", "model": "text-embedding-004", "gen_params": {}, "start": 1732229263.8026, "finish": 1732229264.1403, "ip": "", "conv_id": "c4fe433b9e924218ba5f0ce0e7a634c2", "model_name": "text-embedding-004", "prompt": "I am looking for a paper to help me deal with data constraints as I scale up my language model.", "output": [["I am looking for a paper to help me deal with data constraints as I scale up my language model.", "Title: Scaling Data-Constrained Language Models\n\nAbstract: The current trend of scaling language models involves increasing both parameter count and training dataset size. Extrapolating this trend suggests that training dataset size may soon be limited by the amount of text data available on the internet. Motivated by this limit, we investigate scaling language models in data-constrained regimes. Specifically, we run a large set of experiments varying the extent of data repetition and compute budget, ranging up to 900 billion training tokens and 9 billion parameter models. We find that with constrained data for a fixed compute budget, training with up to 4 epochs of repeated data yields negligible changes to loss compared to having unique data. However, with more repetition, the value of adding compute eventually decays to zero. We propose and empirically validate a scaling law for compute optimality that accounts for the decreasing value of repeated tokens and excess parameters. Finally, we experiment with approaches mitigating data scarcity, including augmenting the training dataset with code data or removing commonly used filters. Models and datasets from our 400 training runs are freely available at https://github.com/huggingface/datablations."]], "corpus": "arxiv"}
+{"tstamp": 1732229264.1403, "task_type": "retrieval", "type": "chat", "model": "embed-english-v3.0", "gen_params": {}, "start": 1732229263.8026, "finish": 1732229264.1403, "ip": "", "conv_id": "fd4a9e8e095e4736bacd093811311cd4", "model_name": "embed-english-v3.0", "prompt": "I am looking for a paper to help me deal with data constraints as I scale up my language model.", "output": [["I am looking for a paper to help me deal with data constraints as I scale up my language model.", "Title: Scaling Data-Constrained Language Models\n\nAbstract: The current trend of scaling language models involves increasing both parameter count and training dataset size. Extrapolating this trend suggests that training dataset size may soon be limited by the amount of text data available on the internet. Motivated by this limit, we investigate scaling language models in data-constrained regimes. Specifically, we run a large set of experiments varying the extent of data repetition and compute budget, ranging up to 900 billion training tokens and 9 billion parameter models. We find that with constrained data for a fixed compute budget, training with up to 4 epochs of repeated data yields negligible changes to loss compared to having unique data. However, with more repetition, the value of adding compute eventually decays to zero. We propose and empirically validate a scaling law for compute optimality that accounts for the decreasing value of repeated tokens and excess parameters. Finally, we experiment with approaches mitigating data scarcity, including augmenting the training dataset with code data or removing commonly used filters. Models and datasets from our 400 training runs are freely available at https://github.com/huggingface/datablations."]], "corpus": "arxiv"}