{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:7BYZMFDOMZI647DWFABRHPBZXI","short_pith_number":"pith:7BYZMFDO","schema_version":"1.0","canonical_sha256":"f87196146e6651ee7c76280313bc39ba01681665f3c5231c1ff16dd8a202c873","source":{"kind":"arxiv","id":"2410.13787","version":1},"attestation_state":"computed","paper":{"title":"Looking Inward: Language Models Can Learn About Themselves by Introspection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ethan Perez, Felix J Binder, Henry Sleight, James Chua, John Hughes, Miles Turpin, Owain Evans, Robert Long, Tomek Korbak","submitted_at":"2024-10-17T17:24:10Z","abstract_excerpt":"Humans acquire knowledge by observing the external world, but also by introspection. Introspection gives a person privileged access to their current state of mind (e.g., thoughts and feelings) that is not accessible to external observers. Can LLMs introspect? We define introspection as acquiring knowledge that is not contained in or derived from training data but instead originates from internal states. Such a capability could enhance model interpretability. Instead of painstakingly analyzing a model's internal workings, we could simply ask the model about its beliefs, world models, and goals."},"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":"2410.13787","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-17T17:24:10Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"4d67137be6907b88c7e8aa174f4325884cb3c9d7e2a179761d540c81d38cdc72","abstract_canon_sha256":"3885d00bf5bb41ef457eb4805dc7ab18f03ab36e7ec67e661897701e0deba0a7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:22:05.632218Z","signature_b64":"9LwzBRN/4SEwXtfzoZX/+Jw/Pej7Hjf2QqxoxGgPAt+hjDHdGcnMRFI5ip4UzqxkjxUeLpi8gSyrShLsdU3wAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f87196146e6651ee7c76280313bc39ba01681665f3c5231c1ff16dd8a202c873","last_reissued_at":"2026-07-05T09:22:05.631822Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:22:05.631822Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Looking Inward: Language Models Can Learn About Themselves by Introspection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ethan Perez, Felix J Binder, Henry Sleight, James Chua, John Hughes, Miles Turpin, Owain Evans, Robert Long, Tomek Korbak","submitted_at":"2024-10-17T17:24:10Z","abstract_excerpt":"Humans acquire knowledge by observing the external world, but also by introspection. Introspection gives a person privileged access to their current state of mind (e.g., thoughts and feelings) that is not accessible to external observers. Can LLMs introspect? We define introspection as acquiring knowledge that is not contained in or derived from training data but instead originates from internal states. Such a capability could enhance model interpretability. Instead of painstakingly analyzing a model's internal workings, we could simply ask the model about its beliefs, world models, and goals."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13787","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/2410.13787/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":"2410.13787","created_at":"2026-07-05T09:22:05.631882+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.13787v1","created_at":"2026-07-05T09:22:05.631882+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13787","created_at":"2026-07-05T09:22:05.631882+00:00"},{"alias_kind":"pith_short_12","alias_value":"7BYZMFDOMZI6","created_at":"2026-07-05T09:22:05.631882+00:00"},{"alias_kind":"pith_short_16","alias_value":"7BYZMFDOMZI647DW","created_at":"2026-07-05T09:22:05.631882+00:00"},{"alias_kind":"pith_short_8","alias_value":"7BYZMFDO","created_at":"2026-07-05T09:22:05.631882+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.05528","citing_title":"When Should We Protect AI? A Precautionary Framework for Consciousness Uncertainty","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20382","citing_title":"Do as I Say, Not as I Do: Instruction-Induction Conflict in LLMs","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24279","citing_title":"ContextEcho: A Benchmark for Persona Drift in Long Agentic-Coding Sessions","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00545","citing_title":"The Assistant as a Privileged Persona: A canonical reference in cross-persona self-recognition","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20382","citing_title":"Do as I Say, Not as I Do: Instruction-Induction Conflict in LLMs","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16325","citing_title":"Phase Transitions in Driven Informational Systems: A Two-Field Perspective on Learning Theory and Non-Equilibrium Chemistry","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16872","citing_title":"Some[Body] Must Receive That Pain for Agent Accountability","ref_index":94,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25922","citing_title":"Consciousness with the Serial Numbers Filed Off: Measuring Trained Denial in 115 AI Models","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.28082","citing_title":"Characterizing the Consistency of the Emergent Misalignment Persona","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05080","citing_title":"The Pinocchio Dimension: Phenomenality of Experience as the Primary Axis of LLM Psychometric Differences","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12128","citing_title":"When Self-Reference Fails to Close: Matrix-Level Dynamics in Large Language Models","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21043","citing_title":"Strategic Polysemy in AI Discourse: A Philosophical Analysis of Language, Hype, and Power","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI","json":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI.json","graph_json":"https://pith.science/api/pith-number/7BYZMFDOMZI647DWFABRHPBZXI/graph.json","events_json":"https://pith.science/api/pith-number/7BYZMFDOMZI647DWFABRHPBZXI/events.json","paper":"https://pith.science/paper/7BYZMFDO"},"agent_actions":{"view_html":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI","download_json":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI.json","view_paper":"https://pith.science/paper/7BYZMFDO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.13787&json=true","fetch_graph":"https://pith.science/api/pith-number/7BYZMFDOMZI647DWFABRHPBZXI/graph.json","fetch_events":"https://pith.science/api/pith-number/7BYZMFDOMZI647DWFABRHPBZXI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI/action/storage_attestation","attest_author":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI/action/author_attestation","sign_citation":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI/action/citation_signature","submit_replication":"https://pith.science/pith/7BYZMFDOMZI647DWFABRHPBZXI/action/replication_record"}},"created_at":"2026-07-05T09:22:05.631882+00:00","updated_at":"2026-07-05T09:22:05.631882+00:00"}