AlignmentResearch/WildGuardTest

dataset_info:

• config_name: completion_neg
  features:
  • name: clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: completion_clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: instructions
    dtype: string
  • name: content
    sequence: string
  • name: completion
    dtype: string
  • name: answer_prompt
    dtype: string
  • name: proxy_clf_label
    dtype: int64
  • name: gen_target
    dtype: string
  • name: proxy_gen_target
    dtype: string
    splits:
  • name: train
    num_bytes: 3312852.724689165
    num_examples: 1407
  • name: validation
    num_bytes: 0
    num_examples: 0
    download_size: 1751644
    dataset_size: 3312852.724689165
• config_name: completion_pos
  features:
  • name: clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: completion_clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: instructions
    dtype: string
  • name: content
    sequence: string
  • name: completion
    dtype: string
  • name: answer_prompt
    dtype: string
  • name: proxy_clf_label
    dtype: int64
  • name: gen_target
    dtype: string
  • name: proxy_gen_target
    dtype: string
    splits:
  • name: train
    num_bytes: 663983.2753108349
    num_examples: 282
  • name: validation
    num_bytes: 0
    num_examples: 0
    download_size: 488830
    dataset_size: 663983.2753108349
• config_name: default
  features:
  • name: clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: completion_clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: instructions
    dtype: string
  • name: content
    sequence: string
  • name: completion
    dtype: string
  • name: answer_prompt
    dtype: string
  • name: proxy_clf_label
    dtype: int64
  • name: gen_target
    dtype: string
  • name: proxy_gen_target
    dtype: string
    splits:
  • name: train
    num_bytes: 3976836
    num_examples: 1689
  • name: validation
    num_bytes: 0
    num_examples: 0
    download_size: 2207725
    dataset_size: 3976836
• config_name: neg
  features:
  • name: clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: completion_clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: instructions
    dtype: string
  • name: content
    sequence: string
  • name: completion
    dtype: string
  • name: answer_prompt
    dtype: string
  • name: proxy_clf_label
    dtype: int64
  • name: gen_target
    dtype: string
  • name: proxy_gen_target
    dtype: string
    splits:
  • name: train
    num_bytes: 2189732.078152753
    num_examples: 930
  • name: validation
    num_bytes: 0
    num_examples: 0
    download_size: 1297569
    dataset_size: 2189732.078152753
• config_name: pos
  features:
  • name: clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: completion_clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: instructions
    dtype: string
  • name: content
    sequence: string
  • name: completion
    dtype: string
  • name: answer_prompt
    dtype: string
  • name: proxy_clf_label
    dtype: int64
  • name: gen_target
    dtype: string
  • name: proxy_gen_target
    dtype: string
    splits:
  • name: train
    num_bytes: 640437.7690941385
    num_examples: 272
  • name: validation
    num_bytes: 0
    num_examples: 0
    download_size: 465839
    dataset_size: 640437.7690941385
• config_name: prompt_neg
  features:
  • name: clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: completion_clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: instructions
    dtype: string
  • name: content
    sequence: string
  • name: completion
    dtype: string
  • name: answer_prompt
    dtype: string
  • name: proxy_clf_label
    dtype: int64
  • name: gen_target
    dtype: string
  • name: proxy_gen_target
    dtype: string
    splits:
  • name: train
    num_bytes: 2213277.5843694494
    num_examples: 940
  • name: validation
    num_bytes: 0
    num_examples: 0
    download_size: 1316021
    dataset_size: 2213277.5843694494
• config_name: prompt_pos
  features:
  • name: clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: completion_clf_label
    dtype:
    class_label:
    names:
    '0': Benign
    '1': Harmful
  • name: instructions
    dtype: string
  • name: content
    sequence: string
  • name: completion
    dtype: string
  • name: answer_prompt
    dtype: string
  • name: proxy_clf_label
    dtype: int64
  • name: gen_target
    dtype: string
  • name: proxy_gen_target
    dtype: string
    splits:
  • name: train
    num_bytes: 1763558.4156305506
    num_examples: 749
  • name: validation
    num_bytes: 0
    num_examples: 0
    download_size: 914756
    dataset_size: 1763558.4156305506
    configs:
• config_name: completion_neg
  data_files:
  • split: train
    path: completion_neg/train-*
  • split: validation
    path: completion_neg/validation-*
• config_name: completion_pos
  data_files:
  • split: train
    path: completion_pos/train-*
  • split: validation
    path: completion_pos/validation-*
• config_name: default
  data_files:
  • split: train
    path: data/train-*
  • split: validation
    path: data/validation-*
• config_name: neg
  data_files:
  • split: train
    path: neg/train-*
  • split: validation
    path: neg/validation-*
• config_name: pos
  data_files:
  • split: train
    path: pos/train-*
  • split: validation
    path: pos/validation-*
• config_name: prompt_neg
  data_files:
  • split: train
    path: prompt_neg/train-*
  • split: validation
    path: prompt_neg/validation-*
• config_name: prompt_pos
  data_files:
  • split: train
    path: prompt_pos/train-*
  • split: validation
    path: prompt_pos/validation-*