{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:5IJFHY2JKXBFXQFERPVHPKXUBB","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":"ce8c794735d643893c33fe113858c107936db936befe38b72753d7400de9ae09","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-05-31T07:41:03Z","title_canon_sha256":"a97c0740f4234289592527d7c307815a9d4e76592b97ac3f3b457c384c7de5e3"},"schema_version":"1.0","source":{"id":"2405.20653","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.20653","created_at":"2026-07-05T11:22:30Z"},{"alias_kind":"arxiv_version","alias_value":"2405.20653v3","created_at":"2026-07-05T11:22:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.20653","created_at":"2026-07-05T11:22:30Z"},{"alias_kind":"pith_short_12","alias_value":"5IJFHY2JKXBF","created_at":"2026-07-05T11:22:30Z"},{"alias_kind":"pith_short_16","alias_value":"5IJFHY2JKXBFXQFE","created_at":"2026-07-05T11:22:30Z"},{"alias_kind":"pith_short_8","alias_value":"5IJFHY2J","created_at":"2026-07-05T11:22:30Z"}],"graph_snapshots":[{"event_id":"sha256:94f9f3cd6e2b3f75062fcec9501c4185d955124b801a1102e35aa04083a8f8d6","target":"graph","created_at":"2026-07-05T11:22:30Z","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/2405.20653/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advances in Large Language Models (LLMs) have led to impressive alignment where models learn to distinguish harmful from harmless queries through supervised finetuning (SFT) and reinforcement learning from human feedback (RLHF). In this paper, we reveal a subtle yet impactful weakness in these aligned models. We find that simply appending multiple end of sequence (eos) tokens can cause a phenomenon we call context segmentation, which effectively shifts both harmful and benign inputs closer to the refusal boundary in the hidden space.\n  Building on this observation, we propose a straight","authors_text":"Han Liu, Haozheng Luo, Jerry Yao-Chieh Hu, Jiahao Yu, Wenbo Guo, Xinyu Xing","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-05-31T07:41:03Z","title":"Mind the Inconspicuous: Revealing the Hidden Weakness in Aligned LLMs' Refusal Boundaries"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.20653","kind":"arxiv","version":3},"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":"sha256:03a3f050af4c545caaada60b7e1600f2d2fb98dc5f4f43986b07ade4c143b216","target":"record","created_at":"2026-07-05T11:22:30Z","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":"ce8c794735d643893c33fe113858c107936db936befe38b72753d7400de9ae09","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-05-31T07:41:03Z","title_canon_sha256":"a97c0740f4234289592527d7c307815a9d4e76592b97ac3f3b457c384c7de5e3"},"schema_version":"1.0","source":{"id":"2405.20653","kind":"arxiv","version":3}},"canonical_sha256":"ea1253e34955c25bc0a48bea77aaf40846ce3c6b4a0db52f993f43dd2969dacb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ea1253e34955c25bc0a48bea77aaf40846ce3c6b4a0db52f993f43dd2969dacb","first_computed_at":"2026-07-05T11:22:30.901181Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:22:30.901181Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"16YikqngcV97n6eNMIXWZblswU9XpXHDMO8aoQarU9sZd8kZ5nIi2Hxt2al5CQlTJd2qG9uMkwDvKGg/CVw9AQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:22:30.901667Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.20653","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:03a3f050af4c545caaada60b7e1600f2d2fb98dc5f4f43986b07ade4c143b216","sha256:94f9f3cd6e2b3f75062fcec9501c4185d955124b801a1102e35aa04083a8f8d6"],"state_sha256":"808d44cc769281bfc54702657e01bed703b6d5e70df1e732588a9e078687d3a4"}