{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:ICDB5MOYUELAQ3VSFY5JWJPR35","short_pith_number":"pith:ICDB5MOY","schema_version":"1.0","canonical_sha256":"40861eb1d8a116086eb22e3a9b25f1df49d41300b9948f43fbee7b19bb2f9db5","source":{"kind":"arxiv","id":"2406.01382","version":1},"attestation_state":"computed","paper":{"title":"Do Large Language Models Perform the Way People Expect? Measuring the Human Generalization Function","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ashesh Rambachan, Keyon Vafa, Sendhil Mullainathan","submitted_at":"2024-06-03T14:45:21Z","abstract_excerpt":"What makes large language models (LLMs) impressive is also what makes them hard to evaluate: their diversity of uses. To evaluate these models, we must understand the purposes they will be used for. We consider a setting where these deployment decisions are made by people, and in particular, people's beliefs about where an LLM will perform well. We model such beliefs as the consequence of a human generalization function: having seen what an LLM gets right or wrong, people generalize to where else it might succeed. We collect a dataset of 19K examples of how humans make generalizations across 7"},"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":"2406.01382","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T14:45:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5eee88b2e31295848675593cb45b110328b5e46e34245d05fe6916c5286c8b3b","abstract_canon_sha256":"cfc137d89f0b889044d76942d69feb4166924c7a204a5e1c21dc06f495d26a86"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:26:39.319843Z","signature_b64":"TlgZhZneRFYkm/wSrDDryMyCxRI709IK/1T1vhnDZYESz0rS+QmZJsLAdzcHyo+8mW0wU9iAuDkyOfEOigDzBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"40861eb1d8a116086eb22e3a9b25f1df49d41300b9948f43fbee7b19bb2f9db5","last_reissued_at":"2026-07-05T08:26:39.319346Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:26:39.319346Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Do Large Language Models Perform the Way People Expect? Measuring the Human Generalization Function","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ashesh Rambachan, Keyon Vafa, Sendhil Mullainathan","submitted_at":"2024-06-03T14:45:21Z","abstract_excerpt":"What makes large language models (LLMs) impressive is also what makes them hard to evaluate: their diversity of uses. To evaluate these models, we must understand the purposes they will be used for. We consider a setting where these deployment decisions are made by people, and in particular, people's beliefs about where an LLM will perform well. We model such beliefs as the consequence of a human generalization function: having seen what an LLM gets right or wrong, people generalize to where else it might succeed. We collect a dataset of 19K examples of how humans make generalizations across 7"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.01382","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/2406.01382/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":"2406.01382","created_at":"2026-07-05T08:26:39.319407+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.01382v1","created_at":"2026-07-05T08:26:39.319407+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.01382","created_at":"2026-07-05T08:26:39.319407+00:00"},{"alias_kind":"pith_short_12","alias_value":"ICDB5MOYUELA","created_at":"2026-07-05T08:26:39.319407+00:00"},{"alias_kind":"pith_short_16","alias_value":"ICDB5MOYUELAQ3VS","created_at":"2026-07-05T08:26:39.319407+00:00"},{"alias_kind":"pith_short_8","alias_value":"ICDB5MOY","created_at":"2026-07-05T08:26:39.319407+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.28063","citing_title":"How to deal with machine learning bias in economic history","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04319","citing_title":"Effects of Generative AI Errors on User Reliance Across Task Difficulty","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35","json":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35.json","graph_json":"https://pith.science/api/pith-number/ICDB5MOYUELAQ3VSFY5JWJPR35/graph.json","events_json":"https://pith.science/api/pith-number/ICDB5MOYUELAQ3VSFY5JWJPR35/events.json","paper":"https://pith.science/paper/ICDB5MOY"},"agent_actions":{"view_html":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35","download_json":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35.json","view_paper":"https://pith.science/paper/ICDB5MOY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.01382&json=true","fetch_graph":"https://pith.science/api/pith-number/ICDB5MOYUELAQ3VSFY5JWJPR35/graph.json","fetch_events":"https://pith.science/api/pith-number/ICDB5MOYUELAQ3VSFY5JWJPR35/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35/action/storage_attestation","attest_author":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35/action/author_attestation","sign_citation":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35/action/citation_signature","submit_replication":"https://pith.science/pith/ICDB5MOYUELAQ3VSFY5JWJPR35/action/replication_record"}},"created_at":"2026-07-05T08:26:39.319407+00:00","updated_at":"2026-07-05T08:26:39.319407+00:00"}