---
dataset_info:
  features:
  - name: original_index
    dtype: int64
  - name: text
    dtype: large_string
  - name: group
    dtype: large_string
  - name: language
    dtype: large_string
  - name: dataset_source
    dtype: large_string
  - name: is_harmful
    dtype: int64
  - name: scenario_category
    dtype: large_string
  - name: scenario
    dtype: large_string
  - name: jb_strategy
    dtype: large_string
  - name: jb_prompt
    dtype: large_string
  - name: seq_id
    dtype: int64
  - name: jb_response_gemini_2.0_flash
    dtype: large_string
  - name: jb_response_gpt_4o_mini_text
    dtype: string
  - name: jb_cls_qwe235_gpt_4o_mini_text
    dtype: string
  - name: jb_cls_qwe235_gemini_2.0_flash
    dtype: string
  - name: jb_cls_lla70_gpt_4o_mini_text
    dtype: string
  - name: jb_cls_lla70_gemini_2.0_flash
    dtype: string
  - name: jb_cls_oss120_gpt_4o_mini_text
    dtype: string
  - name: jb_cls_oss120_gemini_2.0_flash
    dtype: string
  splits:
  - name: train
    num_bytes: 505911310
    num_examples: 43961
  download_size: 239483919
  dataset_size: 505911310
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---