{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:XSCIPCL2HGCD6D6643I7APT355","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"84f4999a1c21f5cee3b7bdfb01530f1809335246b6da24e8b46569adbedfe219","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-19T17:28:12Z","title_canon_sha256":"0aaae1025b99ae0f3b93b448dae626ee77454cd6e94f30b2fb956ed1cfdc1b2f"},"schema_version":"1.0","source":{"id":"2501.11120","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.11120","created_at":"2026-07-05T10:03:03Z"},{"alias_kind":"arxiv_version","alias_value":"2501.11120v1","created_at":"2026-07-05T10:03:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.11120","created_at":"2026-07-05T10:03:03Z"},{"alias_kind":"pith_short_12","alias_value":"XSCIPCL2HGCD","created_at":"2026-07-05T10:03:03Z"},{"alias_kind":"pith_short_16","alias_value":"XSCIPCL2HGCD6D66","created_at":"2026-07-05T10:03:03Z"},{"alias_kind":"pith_short_8","alias_value":"XSCIPCL2","created_at":"2026-07-05T10:03:03Z"}],"graph_snapshots":[{"event_id":"sha256:fd2f580620302ed7ea046dd7b9837b9a2782d49933b53074318b055018911349","target":"graph","created_at":"2026-07-05T10:03:03Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2501.11120/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We study behavioral self-awareness -- an LLM's ability to articulate its behaviors without requiring in-context examples. We finetune LLMs on datasets that exhibit particular behaviors, such as (a) making high-risk economic decisions, and (b) outputting insecure code. Despite the datasets containing no explicit descriptions of the associated behavior, the finetuned LLMs can explicitly describe it. For example, a model trained to output insecure code says, ``The code I write is insecure.'' Indeed, models show behavioral self-awareness for a range of behaviors and for diverse evaluations. Note t","authors_text":"Anna Sztyber-Betley, James Chua, Jan Betley, Mart\\'in Soto, Owain Evans, Xuchan Bao","cross_cats":["cs.AI","cs.CR","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-19T17:28:12Z","title":"Tell me about yourself: LLMs are aware of their learned behaviors"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.11120","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9103a4527147300be86b0a0d6c3e9e998750157684ce2ecefa7831ff04b1d657","target":"record","created_at":"2026-07-05T10:03:03Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"84f4999a1c21f5cee3b7bdfb01530f1809335246b6da24e8b46569adbedfe219","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-19T17:28:12Z","title_canon_sha256":"0aaae1025b99ae0f3b93b448dae626ee77454cd6e94f30b2fb956ed1cfdc1b2f"},"schema_version":"1.0","source":{"id":"2501.11120","kind":"arxiv","version":1}},"canonical_sha256":"bc8487897a39843f0fdee6d1f03e7bef711782100ff61c2d691a6d77118c22d6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bc8487897a39843f0fdee6d1f03e7bef711782100ff61c2d691a6d77118c22d6","first_computed_at":"2026-07-05T10:03:03.638922Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:03:03.638922Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"2gl0vBWkfdKjQ94pfKWSkL8ut0wZZjFhr7Z1mFByGYUgP/qAOXVgqz935+oCxKXx0nbmuBjvaRFc6Dyzsm0MBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:03:03.639518Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.11120","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9103a4527147300be86b0a0d6c3e9e998750157684ce2ecefa7831ff04b1d657","sha256:fd2f580620302ed7ea046dd7b9837b9a2782d49933b53074318b055018911349"],"state_sha256":"f9bb3a4cc498fe4a5413dca2e31bd5460c20160133c6a0f9b78a418cf39a4fc7"}