{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NNSANYQSXZ7XFOJE4HBKUY65Q4","short_pith_number":"pith:NNSANYQS","schema_version":"1.0","canonical_sha256":"6b6406e212be7f72b924e1c2aa63dd871d94484b80c9a108dffeb16f2ebc98e9","source":{"kind":"arxiv","id":"2506.00582","version":2},"attestation_state":"computed","paper":{"title":"Do Language Models Mirror Human Confidence? Exploring Psychological Insights to Address Overconfidence in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bill Howe, Bingbing Wen, Bin Han, Chenjun Xu, Lucy Lu Wang, Robert Wolfe","submitted_at":"2025-05-31T14:37:18Z","abstract_excerpt":"Psychology research has shown that humans are poor at estimating their performance on tasks, tending towards underconfidence on easy tasks and overconfidence on difficult tasks. We examine three LLMs, Llama-3-70B-instruct, Claude-3-Sonnet, and GPT-4o, on a range of QA tasks of varying difficulty, and show that models exhibit subtle differences from human patterns of overconfidence: less sensitive to task difficulty, and when prompted to answer based on different personas -- e.g., expert vs layman, or different race, gender, and ages -- the models will respond with stereotypically biased confid"},"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":"2506.00582","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-31T14:37:18Z","cross_cats_sorted":[],"title_canon_sha256":"4d4a0085ee161afd50907ec72a14e39d7970047d70dbe943c97cf60b39b5cffb","abstract_canon_sha256":"2738d198aeaabe2ee268fe52935fcece6b01382407b4aeeab7fbd7ac5355609c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:06.610227Z","signature_b64":"x3PZOM3cy2c0H9iJcn+7pU9kdORGfHaPKZGr1fIjSKWDHbXTrLMOQG/pRHfqijJrq+e1Z+K3KgQkTyAekZUrDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b6406e212be7f72b924e1c2aa63dd871d94484b80c9a108dffeb16f2ebc98e9","last_reissued_at":"2026-07-05T11:44:06.609492Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:06.609492Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Do Language Models Mirror Human Confidence? Exploring Psychological Insights to Address Overconfidence in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bill Howe, Bingbing Wen, Bin Han, Chenjun Xu, Lucy Lu Wang, Robert Wolfe","submitted_at":"2025-05-31T14:37:18Z","abstract_excerpt":"Psychology research has shown that humans are poor at estimating their performance on tasks, tending towards underconfidence on easy tasks and overconfidence on difficult tasks. We examine three LLMs, Llama-3-70B-instruct, Claude-3-Sonnet, and GPT-4o, on a range of QA tasks of varying difficulty, and show that models exhibit subtle differences from human patterns of overconfidence: less sensitive to task difficulty, and when prompted to answer based on different personas -- e.g., expert vs layman, or different race, gender, and ages -- the models will respond with stereotypically biased confid"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.00582","kind":"arxiv","version":2},"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/2506.00582/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":"2506.00582","created_at":"2026-07-05T11:44:06.609561+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.00582v2","created_at":"2026-07-05T11:44:06.609561+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.00582","created_at":"2026-07-05T11:44:06.609561+00:00"},{"alias_kind":"pith_short_12","alias_value":"NNSANYQSXZ7X","created_at":"2026-07-05T11:44:06.609561+00:00"},{"alias_kind":"pith_short_16","alias_value":"NNSANYQSXZ7XFOJE","created_at":"2026-07-05T11:44:06.609561+00:00"},{"alias_kind":"pith_short_8","alias_value":"NNSANYQS","created_at":"2026-07-05T11:44:06.609561+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.05009","citing_title":"Can Large Language Models Integrate Spatial Data? Empirical Insights into Reasoning Strengths and Computational Weaknesses","ref_index":27,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4","json":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4.json","graph_json":"https://pith.science/api/pith-number/NNSANYQSXZ7XFOJE4HBKUY65Q4/graph.json","events_json":"https://pith.science/api/pith-number/NNSANYQSXZ7XFOJE4HBKUY65Q4/events.json","paper":"https://pith.science/paper/NNSANYQS"},"agent_actions":{"view_html":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4","download_json":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4.json","view_paper":"https://pith.science/paper/NNSANYQS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.00582&json=true","fetch_graph":"https://pith.science/api/pith-number/NNSANYQSXZ7XFOJE4HBKUY65Q4/graph.json","fetch_events":"https://pith.science/api/pith-number/NNSANYQSXZ7XFOJE4HBKUY65Q4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4/action/storage_attestation","attest_author":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4/action/author_attestation","sign_citation":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4/action/citation_signature","submit_replication":"https://pith.science/pith/NNSANYQSXZ7XFOJE4HBKUY65Q4/action/replication_record"}},"created_at":"2026-07-05T11:44:06.609561+00:00","updated_at":"2026-07-05T11:44:06.609561+00:00"}