---
dataset_info:
- config_name: aggregation_cohort
  features:
  - name: cohort
    dtype: int64
  - name: base_period
    dtype: int64
  - name: time
    dtype: int64
  - name: ATT
    dtype: float64
  - name: std_error
    dtype: float64
  - name: lower
    dtype: float64
  - name: upper
    dtype: float64
  - name: zero_not_in_cband
    dtype: string
  - name: model_size
    dtype: string
  - name: stat
    dtype: string
  splits:
  - name: sequence_surprisal
    num_bytes: 12846809
    num_examples: 143143
  - name: sequence_entropy
    num_bytes: 10756208
    num_examples: 122694
  - name: average_rank
    num_bytes: 10266943
    num_examples: 122694
  - name: token_accuracy
    num_bytes: 12270334
    num_examples: 143143
  download_size: 21917713
  dataset_size: 46140294
- config_name: aggregation_event
  features:
  - name: cohort
    dtype: int64
  - name: base_period
    dtype: int64
  - name: time
    dtype: int64
  - name: ATT
    dtype: float64
  - name: std_error
    dtype: float64
  - name: lower
    dtype: float64
  - name: upper
    dtype: float64
  - name: zero_not_in_cband
    dtype: string
  - name: model_size
    dtype: string
  - name: stat
    dtype: string
  splits:
  - name: sequence_surprisal
    num_bytes: 12846809
    num_examples: 143143
  - name: sequence_entropy
    num_bytes: 10756208
    num_examples: 122694
  - name: average_rank
    num_bytes: 10266943
    num_examples: 122694
  - name: token_accuracy
    num_bytes: 12270334
    num_examples: 143143
  download_size: 21917713
  dataset_size: 46140294
- config_name: aggregation_time
  features:
  - name: cohort
    dtype: int64
  - name: base_period
    dtype: int64
  - name: time
    dtype: int64
  - name: ATT
    dtype: float64
  - name: std_error
    dtype: float64
  - name: lower
    dtype: float64
  - name: upper
    dtype: float64
  - name: zero_not_in_cband
    dtype: string
  - name: model_size
    dtype: string
  - name: stat
    dtype: string
  splits:
  - name: sequence_surprisal
    num_bytes: 12846809
    num_examples: 143143
  - name: sequence_entropy
    num_bytes: 10756208
    num_examples: 122694
  - name: average_rank
    num_bytes: 10266943
    num_examples: 122694
  - name: token_accuracy
    num_bytes: 12270334
    num_examples: 143143
  download_size: 21917713
  dataset_size: 46140294
- config_name: default
  features:
  - name: cohort
    dtype: int64
  - name: base_period
    dtype: int64
  - name: time
    dtype: int64
  - name: ATT
    dtype: float64
  - name: std_error
    dtype: float64
  - name: lower
    dtype: float64
  - name: upper
    dtype: float64
  - name: zero_not_in_cband
    dtype: string
  - name: model_size
    dtype: string
  - name: stat
    dtype: string
  splits:
  - name: sequence_surprisal
    num_bytes: 12846809
    num_examples: 143143
  - name: sequence_entropy
    num_bytes: 10756208
    num_examples: 122694
  - name: average_rank
    num_bytes: 10266943
    num_examples: 122694
  - name: token_accuracy
    num_bytes: 12270334
    num_examples: 143143
  download_size: 21917713
  dataset_size: 46140294
configs:
- config_name: aggregation_cohort
  data_files:
  - split: sequence_surprisal
    path: aggregation_cohort/sequence_surprisal-*
  - split: sequence_entropy
    path: aggregation_cohort/sequence_entropy-*
  - split: average_rank
    path: aggregation_cohort/average_rank-*
  - split: token_accuracy
    path: aggregation_cohort/token_accuracy-*
- config_name: aggregation_event
  data_files:
  - split: sequence_surprisal
    path: aggregation_event/sequence_surprisal-*
  - split: sequence_entropy
    path: aggregation_event/sequence_entropy-*
  - split: average_rank
    path: aggregation_event/average_rank-*
  - split: token_accuracy
    path: aggregation_event/token_accuracy-*
- config_name: aggregation_time
  data_files:
  - split: sequence_surprisal
    path: aggregation_time/sequence_surprisal-*
  - split: sequence_entropy
    path: aggregation_time/sequence_entropy-*
  - split: average_rank
    path: aggregation_time/average_rank-*
  - split: token_accuracy
    path: aggregation_time/token_accuracy-*
- config_name: default
  data_files:
  - split: sequence_surprisal
    path: data/sequence_surprisal-*
  - split: sequence_entropy
    path: data/sequence_entropy-*
  - split: average_rank
    path: data/average_rank-*
  - split: token_accuracy
    path: data/token_accuracy-*
---

This dataset has been created as an artefact of the paper [Causal Estimation of Memorisation Profiles (Lesci et al., 2024)](https://arxiv.org/abs/2406.04327). 
More info about this dataset in the related collection [Memorisation-Profiles](https://huggingface.co/collections/pietrolesci/memorisation-profiles-6619604c4594c878cd9d451f).