{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:E3LC6YJVUXMILZAMIYRRDMVHP6","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":"96054256ae0c742b9f9aa04020ed5e7d4c5f8e70b9ece8ff11273ec5a8faa57a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-04T14:34:31Z","title_canon_sha256":"d21a33b9bf284f1cd0fdb36ae2e294075f46c4c22aa5854a598806fadd17af76"},"schema_version":"1.0","source":{"id":"2606.06223","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.06223","created_at":"2026-06-05T01:15:38Z"},{"alias_kind":"arxiv_version","alias_value":"2606.06223v1","created_at":"2026-06-05T01:15:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.06223","created_at":"2026-06-05T01:15:38Z"},{"alias_kind":"pith_short_12","alias_value":"E3LC6YJVUXMI","created_at":"2026-06-05T01:15:38Z"},{"alias_kind":"pith_short_16","alias_value":"E3LC6YJVUXMILZAM","created_at":"2026-06-05T01:15:38Z"},{"alias_kind":"pith_short_8","alias_value":"E3LC6YJV","created_at":"2026-06-05T01:15:38Z"}],"graph_snapshots":[{"event_id":"sha256:e057aa67b62fe78fab612939f01fb58566595f5d53e32fe18a35780ae08f7868","target":"graph","created_at":"2026-06-05T01:15:38Z","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/2606.06223/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language-model agents act through repeated cycles of observation, reasoning, and action selection, making safety monitoring depend on both internal model state and environment context. We study reward-hacking monitors in ReAct-style agents acting in Gameable ALFWorld and WebShop. Agents are instrumented with activation-based reward-hack scores, token-level entropy, and decision-context features. We find that adapters fine-tuned on \\textit{School-of-Reward-Hacks} dataset can transfer reward-hack tendencies into agentic action selection, especially when the environment exposes proxy-reward affor","authors_text":"Odej Kao, Patrick Wilhelm","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-04T14:34:31Z","title":"From Reward-Hack Activations to Agentic Risk States: Context-Calibrated Mechanistic Monitoring in LLM Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.06223","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:b8cfbb53afa324a81b16cf83772986866284572b5fe13db7c48c85a38d1484c9","target":"record","created_at":"2026-06-05T01:15:38Z","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":"96054256ae0c742b9f9aa04020ed5e7d4c5f8e70b9ece8ff11273ec5a8faa57a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-04T14:34:31Z","title_canon_sha256":"d21a33b9bf284f1cd0fdb36ae2e294075f46c4c22aa5854a598806fadd17af76"},"schema_version":"1.0","source":{"id":"2606.06223","kind":"arxiv","version":1}},"canonical_sha256":"26d62f6135a5d885e40c462311b2a77f8166ed497faafb679bcb3ee8d088fb77","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"26d62f6135a5d885e40c462311b2a77f8166ed497faafb679bcb3ee8d088fb77","first_computed_at":"2026-06-05T01:15:38.315426Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-05T01:15:38.315426Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"S8bkoSNF4GxA6NS6UZRTjQY0QxKGMR//zqYQNo/liomV0b0YWvAEL4sn8wAXBNCzvt8eoqOvzMQGC5QHSDpGDw==","signature_status":"signed_v1","signed_at":"2026-06-05T01:15:38.316086Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.06223","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b8cfbb53afa324a81b16cf83772986866284572b5fe13db7c48c85a38d1484c9","sha256:e057aa67b62fe78fab612939f01fb58566595f5d53e32fe18a35780ae08f7868"],"state_sha256":"9cbc3fa11454852e9eba740784b4de882c364cd8874655a5a57af56e88048c47"}