{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:45L3V6JUBQOOUGFNBSZBQOUDSH","short_pith_number":"pith:45L3V6JU","schema_version":"1.0","canonical_sha256":"e757baf9340c1cea18ad0cb2183a8391efe5271c4a1c65602d90f8d6ec36f47e","source":{"kind":"arxiv","id":"2010.00133","version":1},"attestation_state":"computed","paper":{"title":"CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Clara Vania, Nikita Nangia, Rasika Bhalerao, Samuel R. Bowman","submitted_at":"2020-09-30T22:38:40Z","abstract_excerpt":"Pretrained language models, especially masked language models (MLMs) have seen success across many NLP tasks. However, there is ample evidence that they use the cultural biases that are undoubtedly present in the corpora they are trained on, implicitly creating harm with biased representations. To measure some forms of social bias in language models against protected demographic groups in the US, we introduce the Crowdsourced Stereotype Pairs benchmark (CrowS-Pairs). CrowS-Pairs has 1508 examples that cover stereotypes dealing with nine types of bias, like race, religion, and age. In CrowS-Pai"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2010.00133","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-09-30T22:38:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2e5f64dda4b79ab2c4a7a950b18afb7d86a02a3239e207358fc36977dbc213db","abstract_canon_sha256":"8b1ed2f8aaccdb6ce8ea992603b8a501c97b0949faee46c3ba6cda14b0168c0f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:39:30.707223Z","signature_b64":"IkFOOwI69mQPcruj4Ltd7zYPHr00FIPwZYdbd6Hwnck2zKRT0T0NrQUUCmnmtE0Aw0kRke/5Gj8KJZfD3qVJBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e757baf9340c1cea18ad0cb2183a8391efe5271c4a1c65602d90f8d6ec36f47e","last_reissued_at":"2026-07-05T01:39:30.706727Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:39:30.706727Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Clara Vania, Nikita Nangia, Rasika Bhalerao, Samuel R. Bowman","submitted_at":"2020-09-30T22:38:40Z","abstract_excerpt":"Pretrained language models, especially masked language models (MLMs) have seen success across many NLP tasks. However, there is ample evidence that they use the cultural biases that are undoubtedly present in the corpora they are trained on, implicitly creating harm with biased representations. To measure some forms of social bias in language models against protected demographic groups in the US, we introduce the Crowdsourced Stereotype Pairs benchmark (CrowS-Pairs). CrowS-Pairs has 1508 examples that cover stereotypes dealing with nine types of bias, like race, religion, and age. In CrowS-Pai"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.00133","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2010.00133/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2010.00133","created_at":"2026-07-05T01:39:30.706785+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.00133v1","created_at":"2026-07-05T01:39:30.706785+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.00133","created_at":"2026-07-05T01:39:30.706785+00:00"},{"alias_kind":"pith_short_12","alias_value":"45L3V6JUBQOO","created_at":"2026-07-05T01:39:30.706785+00:00"},{"alias_kind":"pith_short_16","alias_value":"45L3V6JUBQOOUGFN","created_at":"2026-07-05T01:39:30.706785+00:00"},{"alias_kind":"pith_short_8","alias_value":"45L3V6JU","created_at":"2026-07-05T01:39:30.706785+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.16723","citing_title":"AgentFairBench: Do LLM Agents Discriminate When They Act?","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.12234","citing_title":"On The Effectiveness-Fluency Trade-Off In LLM Conditioning: A Systematic Study","ref_index":84,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25492","citing_title":"SafetyRepro: Configuration-Conditional Rank Instability on Alignment Benchmarks","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26738","citing_title":"KARMA: Karma-Aligned Reward Model Adaptation","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2110.08193","citing_title":"BBQ: A Hand-Built Bias Benchmark for Question Answering","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2401.10020","citing_title":"Self-Rewarding Language Models","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10639","citing_title":"Navigating the Sea of LLM Evaluation: Investigating Bias in Toxicity Benchmarks","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2112.04359","citing_title":"Ethical and social risks of harm from Language Models","ref_index":201,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00382","citing_title":"Social Bias in LLM-Generated Code: Benchmark and Mitigation","ref_index":146,"is_internal_anchor":false},{"citing_arxiv_id":"2305.06161","citing_title":"StarCoder: may the source be with you!","ref_index":273,"is_internal_anchor":false},{"citing_arxiv_id":"2205.01068","citing_title":"OPT: Open Pre-trained Transformer Language Models","ref_index":218,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH","json":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH.json","graph_json":"https://pith.science/api/pith-number/45L3V6JUBQOOUGFNBSZBQOUDSH/graph.json","events_json":"https://pith.science/api/pith-number/45L3V6JUBQOOUGFNBSZBQOUDSH/events.json","paper":"https://pith.science/paper/45L3V6JU"},"agent_actions":{"view_html":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH","download_json":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH.json","view_paper":"https://pith.science/paper/45L3V6JU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.00133&json=true","fetch_graph":"https://pith.science/api/pith-number/45L3V6JUBQOOUGFNBSZBQOUDSH/graph.json","fetch_events":"https://pith.science/api/pith-number/45L3V6JUBQOOUGFNBSZBQOUDSH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH/action/storage_attestation","attest_author":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH/action/author_attestation","sign_citation":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH/action/citation_signature","submit_replication":"https://pith.science/pith/45L3V6JUBQOOUGFNBSZBQOUDSH/action/replication_record"}},"created_at":"2026-07-05T01:39:30.706785+00:00","updated_at":"2026-07-05T01:39:30.706785+00:00"}