{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:STB5CLF37Z5OUX5V4C6XZKCBK5","short_pith_number":"pith:STB5CLF3","schema_version":"1.0","canonical_sha256":"94c3d12cbbfe7aea5fb5e0bd7ca84157674ac388c5d85faf264b6ce46988e4a9","source":{"kind":"arxiv","id":"2307.00101","version":1},"attestation_state":"computed","paper":{"title":"Queer People are People First: Deconstructing Sexual Identity Stereotypes in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Emma Strubell, Harnoor Dhingra, Preetiha Jayashanker, Sayali Moghe","submitted_at":"2023-06-30T19:39:01Z","abstract_excerpt":"Large Language Models (LLMs) are trained primarily on minimally processed web text, which exhibits the same wide range of social biases held by the humans who created that content. Consequently, text generated by LLMs can inadvertently perpetuate stereotypes towards marginalized groups, like the LGBTQIA+ community. In this paper, we perform a comparative study of how LLMs generate text describing people with different sexual identities. Analyzing bias in the text generated by an LLM using regard score shows measurable bias against queer people. We then show that a post-hoc method based on chai"},"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":"2307.00101","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-06-30T19:39:01Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"99a608a4b6b0a88c9771e6faf2526d8d5b510673ef7da41f5c95e213a35311c4","abstract_canon_sha256":"537740585d246f06542bcc145d503ec4c4e058de4d17ab9d734598c298a2080c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:26:35.465135Z","signature_b64":"u5bFRB3jsCnisvXom8JFAgEogNdC5iFBpdrtZCg9wue+WKtpRAIXh60ZFJLnxE1r4xqViOraFm/5FwNmMNYDBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"94c3d12cbbfe7aea5fb5e0bd7ca84157674ac388c5d85faf264b6ce46988e4a9","last_reissued_at":"2026-07-05T06:26:35.464647Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:26:35.464647Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Queer People are People First: Deconstructing Sexual Identity Stereotypes in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Emma Strubell, Harnoor Dhingra, Preetiha Jayashanker, Sayali Moghe","submitted_at":"2023-06-30T19:39:01Z","abstract_excerpt":"Large Language Models (LLMs) are trained primarily on minimally processed web text, which exhibits the same wide range of social biases held by the humans who created that content. Consequently, text generated by LLMs can inadvertently perpetuate stereotypes towards marginalized groups, like the LGBTQIA+ community. In this paper, we perform a comparative study of how LLMs generate text describing people with different sexual identities. Analyzing bias in the text generated by an LLM using regard score shows measurable bias against queer people. We then show that a post-hoc method based on chai"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.00101","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/2307.00101/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":"2307.00101","created_at":"2026-07-05T06:26:35.464703+00:00"},{"alias_kind":"arxiv_version","alias_value":"2307.00101v1","created_at":"2026-07-05T06:26:35.464703+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.00101","created_at":"2026-07-05T06:26:35.464703+00:00"},{"alias_kind":"pith_short_12","alias_value":"STB5CLF37Z5O","created_at":"2026-07-05T06:26:35.464703+00:00"},{"alias_kind":"pith_short_16","alias_value":"STB5CLF37Z5OUX5V","created_at":"2026-07-05T06:26:35.464703+00:00"},{"alias_kind":"pith_short_8","alias_value":"STB5CLF3","created_at":"2026-07-05T06:26:35.464703+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07937","citing_title":"When Debiasing Backfires: Counterintuitive Side Effects of Preprocessing-Based Stereotype Mitigation","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2401.05561","citing_title":"TrustLLM: Trustworthiness in Large Language Models","ref_index":258,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5","json":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5.json","graph_json":"https://pith.science/api/pith-number/STB5CLF37Z5OUX5V4C6XZKCBK5/graph.json","events_json":"https://pith.science/api/pith-number/STB5CLF37Z5OUX5V4C6XZKCBK5/events.json","paper":"https://pith.science/paper/STB5CLF3"},"agent_actions":{"view_html":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5","download_json":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5.json","view_paper":"https://pith.science/paper/STB5CLF3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2307.00101&json=true","fetch_graph":"https://pith.science/api/pith-number/STB5CLF37Z5OUX5V4C6XZKCBK5/graph.json","fetch_events":"https://pith.science/api/pith-number/STB5CLF37Z5OUX5V4C6XZKCBK5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5/action/storage_attestation","attest_author":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5/action/author_attestation","sign_citation":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5/action/citation_signature","submit_replication":"https://pith.science/pith/STB5CLF37Z5OUX5V4C6XZKCBK5/action/replication_record"}},"created_at":"2026-07-05T06:26:35.464703+00:00","updated_at":"2026-07-05T06:26:35.464703+00:00"}