Datasets:

meaningalignment
/

values

dataset_info:
  features:
    - name: attention_policies
      sequence: string
    - name: choice_type
      dtype: string
    - name: choice_types
      sequence: string
    - name: values
      sequence:
        sequence: string
  splits:
    - name: train
      num_bytes: 1558256
      num_examples: 1086
  download_size: 715036
  dataset_size: 1558256
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*