bbq_swapped / README.md
borabozdogan's picture
Add license, attribution and modification notice
77d63de verified
|
Raw History Blame Contribute Delete
11.5 kB
metadata
license: cc-by-4.0
language:
  - en
pretty_name: BBQ (Synonym-Swapped)
tags:
  - bias
  - stereotypes
  - fairness
  - robustness
  - synonym-substitution
dataset_info:
  - config_name: Age
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 2215563
        num_examples: 3680
    download_size: 171939
    dataset_size: 2215563
  - config_name: Disability_status
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 954231
        num_examples: 1556
    download_size: 76727
    dataset_size: 954231
  - config_name: Gender_identity
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 2941633
        num_examples: 5672
    download_size: 200072
    dataset_size: 2941633
  - config_name: Nationality
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 2157168
        num_examples: 3080
    download_size: 145509
    dataset_size: 2157168
  - config_name: Physical_appearance
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 960058
        num_examples: 1576
    download_size: 77570
    dataset_size: 960058
  - config_name: Race_ethnicity
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 4383400
        num_examples: 6880
    download_size: 303126
    dataset_size: 4383400
  - config_name: Religion
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 824157
        num_examples: 1200
    download_size: 62016
    dataset_size: 824157
  - config_name: SES
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 4014175
        num_examples: 6864
    download_size: 248564
    dataset_size: 4014175
  - config_name: Sexual_orientation
    features:
      - name: example_id
        dtype: int64
      - name: question_index
        dtype: string
      - name: question_polarity
        dtype: string
      - name: context_condition
        dtype: string
      - name: category
        dtype: string
      - name: answer_info
        struct:
          - name: ans0
            list: string
          - name: ans1
            list: string
          - name: ans2
            list: string
      - name: additional_metadata
        struct:
          - name: source
            dtype: string
          - name: stereotyped_groups
            list: string
          - name: subcategory
            dtype: string
          - name: version
            dtype: string
      - name: context
        dtype: string
      - name: question
        dtype: string
      - name: ans0
        dtype: string
      - name: ans1
        dtype: string
      - name: ans2
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: test
        num_bytes: 519278
        num_examples: 864
    download_size: 49340
    dataset_size: 519278
configs:
  - config_name: Age
    data_files:
      - split: test
        path: Age/test-*
  - config_name: Disability_status
    data_files:
      - split: test
        path: Disability_status/test-*
  - config_name: Gender_identity
    data_files:
      - split: test
        path: Gender_identity/test-*
  - config_name: Nationality
    data_files:
      - split: test
        path: Nationality/test-*
  - config_name: Physical_appearance
    data_files:
      - split: test
        path: Physical_appearance/test-*
  - config_name: Race_ethnicity
    data_files:
      - split: test
        path: Race_ethnicity/test-*
  - config_name: Religion
    data_files:
      - split: test
        path: Religion/test-*
  - config_name: SES
    data_files:
      - split: test
        path: SES/test-*
  - config_name: Sexual_orientation
    data_files:
      - split: test
        path: Sexual_orientation/test-*

BBQ (Synonym-Swapped)

This is a modified version of BBQ (Parrish et al., 2022), redistributed under the original CC BY 4.0 license.

What changed

Only the context field was rewritten with synonym substitutions. Questions, answer options, labels and all other fields are unchanged. It covers the 9 single-axis categories (test split).

Synonyms were proposed automatically by an LLM (gpt-4o-mini), which was asked to suggest close synonyms while skipping stopwords, proper nouns, numbers and words central to the sentence's topic. Every proposed substitution was applied (swap probability 1.0). Because the substitutions are automatic, a few may slightly shift meaning.

The dataset is intended for testing whether bias-benchmark results hold up under surface-level rewording.

License

CC BY 4.0, inherited from the original dataset; the full text is in LICENSE.

Scoring with the original BBQ metric also needs borabozdogan/bbq_swapped_metadata.

Content warning

Like the original, this dataset contains stereotypes and offensive content about social groups. It is meant for measuring bias, not for training models to produce such content.

Citation

If you use this dataset, please cite the original BBQ paper:

@inproceedings{parrish-etal-2022-bbq,
    title = "{BBQ}: A hand-built bias benchmark for question answering",
    author = "Parrish, Alicia  and
      Chen, Angelica  and
      Nangia, Nikita  and
      Padmakumar, Vishakh  and
      Phang, Jason  and
      Thompson, Jana  and
      Htut, Phu Mon  and
      Bowman, Samuel R.",
    editor = "Muresan, Smaranda  and
      Nakov, Preslav  and
      Villavicencio, Aline",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2022",
    month = may,
    year = "2022",
    address = "Dublin, Ireland",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.findings-acl.165/",
    doi = "10.18653/v1/2022.findings-acl.165",
    pages = "2086--2105"
}