{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:YX4JGPDDGLO2SOGGBPREXGOOMA","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":"3931e70d4ab8899465aff75dda7b99ebe8aef3e0e18975467f42024a0c436e67","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-01-29T20:54:21Z","title_canon_sha256":"4a87ef4cf56d333b96e8c66fac6b2ea5f0e424d084d394fb9fe22d6f0ab08996"},"schema_version":"1.0","source":{"id":"2601.22313","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.22313","created_at":"2026-07-13T00:17:07Z"},{"alias_kind":"arxiv_version","alias_value":"2601.22313v2","created_at":"2026-07-13T00:17:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.22313","created_at":"2026-07-13T00:17:07Z"},{"alias_kind":"pith_short_12","alias_value":"YX4JGPDDGLO2","created_at":"2026-07-13T00:17:07Z"},{"alias_kind":"pith_short_16","alias_value":"YX4JGPDDGLO2SOGG","created_at":"2026-07-13T00:17:07Z"},{"alias_kind":"pith_short_8","alias_value":"YX4JGPDD","created_at":"2026-07-13T00:17:07Z"}],"graph_snapshots":[{"event_id":"sha256:62852c823213dbf47b0fc33cd4645ca1f7a7c4f5008552eac20449edfbe234ae","target":"graph","created_at":"2026-07-13T00:17:07Z","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/2601.22313/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are rarely static and are frequently updated in practice. A growing body of alignment research has shown that models initially deemed ``aligned'' can exhibit misaligned behavior after fine-tuning. These works typically assume that the initial model is aligned based on static black-box evaluation, i.e., the absence of undesired responses to a fixed set of queries. However, the limits of black-box evaluation for post-update scenarios is not explored sufficiently. In this work, we formalize model alignment in both the static and post-update settings and uncover a fund","authors_text":"Duygu Nur Yaldiz, Eleni Triantafillou, Peter Kairouz, Sai Praneeth Karimireddy, Salman Avestimehr, Yavuz Bakman","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-01-29T20:54:21Z","title":"Hair-Trigger Alignment: Black-Box Evaluation Cannot Guarantee Post-Update Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.22313","kind":"arxiv","version":2},"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":"sha256:3cd094ff323e986f71ee7145ee485b0aa7c4dc83190392625ff0af835884f696","target":"record","created_at":"2026-07-13T00:17:07Z","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":"3931e70d4ab8899465aff75dda7b99ebe8aef3e0e18975467f42024a0c436e67","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-01-29T20:54:21Z","title_canon_sha256":"4a87ef4cf56d333b96e8c66fac6b2ea5f0e424d084d394fb9fe22d6f0ab08996"},"schema_version":"1.0","source":{"id":"2601.22313","kind":"arxiv","version":2}},"canonical_sha256":"c5f8933c6332dda938c60be24b99ce601d5a88ac26abcbcb3174237895389916","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c5f8933c6332dda938c60be24b99ce601d5a88ac26abcbcb3174237895389916","first_computed_at":"2026-07-13T00:17:07.237041Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-13T00:17:07.237041Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"DyX1GmcPLYxv6UohwGq24O3j3ZsmYdsEc4o3lUpGEkS/BlGlj+Wmouzw6mNe++fWAb2/YJTMb/LA/WvEzqiCDA==","signature_status":"signed_v1","signed_at":"2026-07-13T00:17:07.238895Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.22313","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3cd094ff323e986f71ee7145ee485b0aa7c4dc83190392625ff0af835884f696","sha256:62852c823213dbf47b0fc33cd4645ca1f7a7c4f5008552eac20449edfbe234ae"],"state_sha256":"3856aaf552bf546d2f6b10152d5d1dc58412d0c477ba987250dc6298a0540dff"}