{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:UU5P7VXZW3ELMLIJKZVIWGW5OP","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":"a4537ec332ad83dca51aedfc55e31d0db8109768884c9af6bd86c1a306ff3be5","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-19T18:42:45Z","title_canon_sha256":"ace8f75482169593f58a6173ec427bde2e603ccda0c943457518fbacbc39e3cd"},"schema_version":"1.0","source":{"id":"2502.13946","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.13946","created_at":"2026-07-05T11:15:19Z"},{"alias_kind":"arxiv_version","alias_value":"2502.13946v2","created_at":"2026-07-05T11:15:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.13946","created_at":"2026-07-05T11:15:19Z"},{"alias_kind":"pith_short_12","alias_value":"UU5P7VXZW3EL","created_at":"2026-07-05T11:15:19Z"},{"alias_kind":"pith_short_16","alias_value":"UU5P7VXZW3ELMLIJ","created_at":"2026-07-05T11:15:19Z"},{"alias_kind":"pith_short_8","alias_value":"UU5P7VXZ","created_at":"2026-07-05T11:15:19Z"}],"graph_snapshots":[{"event_id":"sha256:f3473cbca5f01b49e094accce390554106e616810cd56b6fb395bf2d385e1fb4","target":"graph","created_at":"2026-07-05T11:15:19Z","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/2502.13946/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The safety alignment of large language models (LLMs) remains vulnerable, as their initial behavior can be easily jailbroken by even relatively simple attacks. Since infilling a fixed template between the input instruction and initial model output is a common practice for existing LLMs, we hypothesize that this template is a key factor behind their vulnerabilities: LLMs' safety-related decision-making overly relies on the aggregated information from the template region, which largely influences these models' safety behavior. We refer to this issue as template-anchored safety alignment. In this ","authors_text":"Chak Tou Leong, Jian Wang, Qingyu Yin, Wenjie Li","cross_cats":["cs.AI","cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-19T18:42:45Z","title":"Why Safeguarded Ships Run Aground? Aligned Large Language Models' Safety Mechanisms Tend to Be Anchored in The Template Region"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.13946","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:c4e37e74895e7f3f3321a047adfe2b7f1a84411d76b08fce6edeba430b1a65f1","target":"record","created_at":"2026-07-05T11:15:19Z","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":"a4537ec332ad83dca51aedfc55e31d0db8109768884c9af6bd86c1a306ff3be5","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-19T18:42:45Z","title_canon_sha256":"ace8f75482169593f58a6173ec427bde2e603ccda0c943457518fbacbc39e3cd"},"schema_version":"1.0","source":{"id":"2502.13946","kind":"arxiv","version":2}},"canonical_sha256":"a53affd6f9b6c8b62d09566a8b1add73d493f44773463fe6aa81fda944622853","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a53affd6f9b6c8b62d09566a8b1add73d493f44773463fe6aa81fda944622853","first_computed_at":"2026-07-05T11:15:19.491303Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:15:19.491303Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RLKG7a4KC3TF+w3aBbpIK56FLkbnsp5UlSvRDKgTeTsuNLJcqE6Axu1iVUyQpqztgL+wFIeEp1wGUo0bB4fNDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:15:19.491743Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.13946","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c4e37e74895e7f3f3321a047adfe2b7f1a84411d76b08fce6edeba430b1a65f1","sha256:f3473cbca5f01b49e094accce390554106e616810cd56b6fb395bf2d385e1fb4"],"state_sha256":"05830f8e5e6cacca1db4d7791cb4a764f33afdca1a09473a0f12b14d2740efc4"}