{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:GIP4KOJYTMRBVB267LFIGPGUQ2","short_pith_number":"pith:GIP4KOJY","schema_version":"1.0","canonical_sha256":"321fc539389b221a875efaca833cd4868b8835eb0933a5d7153c1488f58be4d0","source":{"kind":"arxiv","id":"2010.02428","version":3},"attestation_state":"computed","paper":{"title":"UnQovering Stereotyping Biases via Underspecified Questions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ashish Sabharwal, Daniel Khashabi, Tao Li, Tushar Khot, Vivek Srikumar","submitted_at":"2020-10-06T01:49:52Z","abstract_excerpt":"While language embeddings have been shown to have stereotyping biases, how these biases affect downstream question answering (QA) models remains unexplored. We present UNQOVER, a general framework to probe and quantify biases through underspecified questions. We show that a naive use of model scores can lead to incorrect bias estimates due to two forms of reasoning errors: positional dependence and question independence. We design a formalism that isolates the aforementioned errors. As case studies, we use this metric to analyze four important classes of stereotypes: gender, nationality, ethni"},"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":"2010.02428","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-10-06T01:49:52Z","cross_cats_sorted":[],"title_canon_sha256":"09e2516babba8a83a2f3c30ed4c1341ca2682cfa56d96c05103f6fb086ca1120","abstract_canon_sha256":"54a52593bab1e875fe2655576202b0feb1f5db87b71afc3a833badb6c4e968b1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:41:58.013641Z","signature_b64":"uF1Z9XFkGogFqnFAxZcYVxE7rgQNLZJHTPDhDcvmKAuZKzaMxbahyLGXpgh8bRt/xwnQkcuBLBirlNOoWdYaAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"321fc539389b221a875efaca833cd4868b8835eb0933a5d7153c1488f58be4d0","last_reissued_at":"2026-07-05T01:41:58.013189Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:41:58.013189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"UnQovering Stereotyping Biases via Underspecified Questions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ashish Sabharwal, Daniel Khashabi, Tao Li, Tushar Khot, Vivek Srikumar","submitted_at":"2020-10-06T01:49:52Z","abstract_excerpt":"While language embeddings have been shown to have stereotyping biases, how these biases affect downstream question answering (QA) models remains unexplored. We present UNQOVER, a general framework to probe and quantify biases through underspecified questions. We show that a naive use of model scores can lead to incorrect bias estimates due to two forms of reasoning errors: positional dependence and question independence. We design a formalism that isolates the aforementioned errors. As case studies, we use this metric to analyze four important classes of stereotypes: gender, nationality, ethni"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.02428","kind":"arxiv","version":3},"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/2010.02428/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":"2010.02428","created_at":"2026-07-05T01:41:58.013245+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.02428v3","created_at":"2026-07-05T01:41:58.013245+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.02428","created_at":"2026-07-05T01:41:58.013245+00:00"},{"alias_kind":"pith_short_12","alias_value":"GIP4KOJYTMRB","created_at":"2026-07-05T01:41:58.013245+00:00"},{"alias_kind":"pith_short_16","alias_value":"GIP4KOJYTMRBVB26","created_at":"2026-07-05T01:41:58.013245+00:00"},{"alias_kind":"pith_short_8","alias_value":"GIP4KOJY","created_at":"2026-07-05T01:41:58.013245+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.23111","citing_title":"FairI Tales: Evaluation of Fairness in Indian Contexts with a Focus on Bias and Stereotypes","ref_index":36,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2","json":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2.json","graph_json":"https://pith.science/api/pith-number/GIP4KOJYTMRBVB267LFIGPGUQ2/graph.json","events_json":"https://pith.science/api/pith-number/GIP4KOJYTMRBVB267LFIGPGUQ2/events.json","paper":"https://pith.science/paper/GIP4KOJY"},"agent_actions":{"view_html":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2","download_json":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2.json","view_paper":"https://pith.science/paper/GIP4KOJY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.02428&json=true","fetch_graph":"https://pith.science/api/pith-number/GIP4KOJYTMRBVB267LFIGPGUQ2/graph.json","fetch_events":"https://pith.science/api/pith-number/GIP4KOJYTMRBVB267LFIGPGUQ2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2/action/storage_attestation","attest_author":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2/action/author_attestation","sign_citation":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2/action/citation_signature","submit_replication":"https://pith.science/pith/GIP4KOJYTMRBVB267LFIGPGUQ2/action/replication_record"}},"created_at":"2026-07-05T01:41:58.013245+00:00","updated_at":"2026-07-05T01:41:58.013245+00:00"}