{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:TI6HHS5G5FP4ARX2SNS7G3QZYS","short_pith_number":"pith:TI6HHS5G","schema_version":"1.0","canonical_sha256":"9a3c73cba6e95fc046fa9365f36e19c4b8dce32c2eb2458a433435743c2b8a32","source":{"kind":"arxiv","id":"2305.18189","version":1},"attestation_state":"computed","paper":{"title":"Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CY"],"primary_cat":"cs.CL","authors_text":"Dan Jurafsky, Esin Durmus, Myra Cheng","submitted_at":"2023-05-29T16:29:22Z","abstract_excerpt":"To recognize and mitigate harms from large language models (LLMs), we need to understand the prevalence and nuances of stereotypes in LLM outputs. Toward this end, we present Marked Personas, a prompt-based method to measure stereotypes in LLMs for intersectional demographic groups without any lexicon or data labeling. Grounded in the sociolinguistic concept of markedness (which characterizes explicitly linguistically marked categories versus unmarked defaults), our proposed method is twofold: 1) prompting an LLM to generate personas, i.e., natural language descriptions, of the target demograp"},"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":"2305.18189","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-29T16:29:22Z","cross_cats_sorted":["cs.AI","cs.CY"],"title_canon_sha256":"1c26b4f89124e95cdb6304a88e112ca2e569d7304f5e1d1b7e53488464566d19","abstract_canon_sha256":"75de2b76a34766d1c0035d80f248c83ac362ed2cde61a6a44db3babcfb3efc01"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:14:53.236574Z","signature_b64":"WXcFdeQOZeONPl35UIY/L74GMJnGSqHPNHJvBpm4N52sxi2jg+IBAj1cuytDK+j44RkL7jnw9Xcutm0L5faBAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a3c73cba6e95fc046fa9365f36e19c4b8dce32c2eb2458a433435743c2b8a32","last_reissued_at":"2026-07-05T06:14:53.236192Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:14:53.236192Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CY"],"primary_cat":"cs.CL","authors_text":"Dan Jurafsky, Esin Durmus, Myra Cheng","submitted_at":"2023-05-29T16:29:22Z","abstract_excerpt":"To recognize and mitigate harms from large language models (LLMs), we need to understand the prevalence and nuances of stereotypes in LLM outputs. Toward this end, we present Marked Personas, a prompt-based method to measure stereotypes in LLMs for intersectional demographic groups without any lexicon or data labeling. Grounded in the sociolinguistic concept of markedness (which characterizes explicitly linguistically marked categories versus unmarked defaults), our proposed method is twofold: 1) prompting an LLM to generate personas, i.e., natural language descriptions, of the target demograp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.18189","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/2305.18189/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":"2305.18189","created_at":"2026-07-05T06:14:53.236245+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.18189v1","created_at":"2026-07-05T06:14:53.236245+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.18189","created_at":"2026-07-05T06:14:53.236245+00:00"},{"alias_kind":"pith_short_12","alias_value":"TI6HHS5G5FP4","created_at":"2026-07-05T06:14:53.236245+00:00"},{"alias_kind":"pith_short_16","alias_value":"TI6HHS5G5FP4ARX2","created_at":"2026-07-05T06:14:53.236245+00:00"},{"alias_kind":"pith_short_8","alias_value":"TI6HHS5G","created_at":"2026-07-05T06:14:53.236245+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.00937","citing_title":"Persona Non Grata: LLM Persona-Driven Generations in MCQA are Unstable in Distinct Dimensions","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15217","citing_title":"Fair outputs, Biased Internals: Causal Potency and Asymmetry of Latent Bias in LLMs for High-Stakes Decisions","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06196","citing_title":"The Granularity Axis: A Micro-to-Macro Latent Direction for Social Roles in Language Models","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07896","citing_title":"What if AI systems weren't chatbots?","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20131","citing_title":"Whose Story Gets Told? Positionality and Bias in LLM Summaries of Life Narratives","ref_index":289,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS","json":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS.json","graph_json":"https://pith.science/api/pith-number/TI6HHS5G5FP4ARX2SNS7G3QZYS/graph.json","events_json":"https://pith.science/api/pith-number/TI6HHS5G5FP4ARX2SNS7G3QZYS/events.json","paper":"https://pith.science/paper/TI6HHS5G"},"agent_actions":{"view_html":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS","download_json":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS.json","view_paper":"https://pith.science/paper/TI6HHS5G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.18189&json=true","fetch_graph":"https://pith.science/api/pith-number/TI6HHS5G5FP4ARX2SNS7G3QZYS/graph.json","fetch_events":"https://pith.science/api/pith-number/TI6HHS5G5FP4ARX2SNS7G3QZYS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS/action/storage_attestation","attest_author":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS/action/author_attestation","sign_citation":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS/action/citation_signature","submit_replication":"https://pith.science/pith/TI6HHS5G5FP4ARX2SNS7G3QZYS/action/replication_record"}},"created_at":"2026-07-05T06:14:53.236245+00:00","updated_at":"2026-07-05T06:14:53.236245+00:00"}