Datasets:

maximuspowers
/

llm-interpretability-v1

dataset_info:
  features:
    - name: text
      dtype: string
    - name: prompt
      dtype: string
    - name: completion
      dtype: string
    - name: metadata
      dtype: string
    - name: example_id
      dtype: int64
  splits:
    - name: train
      num_bytes: 395207764
      num_examples: 1000
  download_size: 206233536
  dataset_size: 395207764
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*