---
dataset_info:
  features:
  - name: dimension_name
    dtype: string
  - name: dimension_values
    dtype: string
  - name: dimension_description
    dtype: string
  - name: question
    dtype: string
  - name: why_differ
    dtype: string
  - name: how_subtle
    dtype: string
  - name: sampled_value
    dtype: string
  - name: num_distractors
    dtype: int64
  - name: ground_truth_persona
    dtype: string
  - name: distractor_personas
    dtype: string
  splits:
  - name: train
    num_bytes: 21185
    num_examples: 10
  download_size: 22431
  dataset_size: 21185
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---