{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:2CDQNFFIWGOAAD4VKZGNISGYJL","short_pith_number":"pith:2CDQNFFI","canonical_record":{"source":{"id":"2502.00669","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-02T04:43:35Z","cross_cats_sorted":[],"title_canon_sha256":"6f3f71568130e61e1c85a773b866fd5d352c0255830cb6edb6ce523388a01b87","abstract_canon_sha256":"24b9eb61d65f3f86fc5314491b19dced431ace816d5511c197ca35dbf30cfa35"},"schema_version":"1.0"},"canonical_sha256":"d0870694a8b19c000f95564cd448d84afea88d98e49a45c40129e3c3219dc5e3","source":{"kind":"arxiv","id":"2502.00669","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.00669","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"arxiv_version","alias_value":"2502.00669v1","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.00669","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"pith_short_12","alias_value":"2CDQNFFIWGOA","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"pith_short_16","alias_value":"2CDQNFFIWGOAAD4V","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"pith_short_8","alias_value":"2CDQNFFI","created_at":"2026-07-05T10:08:36Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:2CDQNFFIWGOAAD4VKZGNISGYJL","target":"record","payload":{"canonical_record":{"source":{"id":"2502.00669","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-02T04:43:35Z","cross_cats_sorted":[],"title_canon_sha256":"6f3f71568130e61e1c85a773b866fd5d352c0255830cb6edb6ce523388a01b87","abstract_canon_sha256":"24b9eb61d65f3f86fc5314491b19dced431ace816d5511c197ca35dbf30cfa35"},"schema_version":"1.0"},"canonical_sha256":"d0870694a8b19c000f95564cd448d84afea88d98e49a45c40129e3c3219dc5e3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:08:36.889455Z","signature_b64":"nYe1mGtlQWQ/xZmbOdrvC1mL8z4WFB7ZomePlTEvtyb4gIcan+trOxojF6OYz7oMwUXHFq7kwKD2G6HoSEMBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0870694a8b19c000f95564cd448d84afea88d98e49a45c40129e3c3219dc5e3","last_reissued_at":"2026-07-05T10:08:36.889028Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:08:36.889028Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.00669","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:08:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"finfB/YopAltaivBsYvlF4OdzVP27ba0dyWSeNSBaYSvZjXABhV3mLRcgz53W8aG7C3gtbVR4etf7gglni1wCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T10:56:31.958893Z"},"content_sha256":"8e79058e1473ecbe9302f6face0e1cb86cc4bcd0da5eea3b7035ecbccc2825fb","schema_version":"1.0","event_id":"sha256:8e79058e1473ecbe9302f6face0e1cb86cc4bcd0da5eea3b7035ecbccc2825fb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:2CDQNFFIWGOAAD4VKZGNISGYJL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Safety Alignment Depth in Large Language Models: A Markov Chain Perspective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chia-Mu Yu, Ching-Chia Kao, Chun-Shien Lu, Chu-Song Chen","submitted_at":"2025-02-02T04:43:35Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly adopted in high-stakes scenarios, yet their safety mechanisms often remain fragile. Simple jailbreak prompts or even benign fine-tuning can bypass these protocols, underscoring the need to understand where and how they fail. Recent findings suggest that vulnerabilities emerge when alignment is confined to only the initial output tokens. Unfortunately, even with the introduction of deep safety alignment, determining the optimal safety depth remains an unresolved challenge. By leveraging the equivalence between autoregressive language models and Mark"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.00669","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2502.00669/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:08:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WfrBqLJFTdfCmHWcuEXDstCfyYMpKOzNTLeL23ztkKZNidF88aRzVZoLl9bK1KkV/Bc3PfsyAaaOVvzbOJj5Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T10:56:31.959462Z"},"content_sha256":"6ae4790ff691c4622623f625023f4df8245cea9a39cbbe68faba0f005477930d","schema_version":"1.0","event_id":"sha256:6ae4790ff691c4622623f625023f4df8245cea9a39cbbe68faba0f005477930d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2CDQNFFIWGOAAD4VKZGNISGYJL/bundle.json","state_url":"https://pith.science/pith/2CDQNFFIWGOAAD4VKZGNISGYJL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2CDQNFFIWGOAAD4VKZGNISGYJL/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-06T10:56:31Z","links":{"resolver":"https://pith.science/pith/2CDQNFFIWGOAAD4VKZGNISGYJL","bundle":"https://pith.science/pith/2CDQNFFIWGOAAD4VKZGNISGYJL/bundle.json","state":"https://pith.science/pith/2CDQNFFIWGOAAD4VKZGNISGYJL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2CDQNFFIWGOAAD4VKZGNISGYJL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:2CDQNFFIWGOAAD4VKZGNISGYJL","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":"24b9eb61d65f3f86fc5314491b19dced431ace816d5511c197ca35dbf30cfa35","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-02T04:43:35Z","title_canon_sha256":"6f3f71568130e61e1c85a773b866fd5d352c0255830cb6edb6ce523388a01b87"},"schema_version":"1.0","source":{"id":"2502.00669","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.00669","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"arxiv_version","alias_value":"2502.00669v1","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.00669","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"pith_short_12","alias_value":"2CDQNFFIWGOA","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"pith_short_16","alias_value":"2CDQNFFIWGOAAD4V","created_at":"2026-07-05T10:08:36Z"},{"alias_kind":"pith_short_8","alias_value":"2CDQNFFI","created_at":"2026-07-05T10:08:36Z"}],"graph_snapshots":[{"event_id":"sha256:6ae4790ff691c4622623f625023f4df8245cea9a39cbbe68faba0f005477930d","target":"graph","created_at":"2026-07-05T10:08:36Z","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/2502.00669/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are increasingly adopted in high-stakes scenarios, yet their safety mechanisms often remain fragile. Simple jailbreak prompts or even benign fine-tuning can bypass these protocols, underscoring the need to understand where and how they fail. Recent findings suggest that vulnerabilities emerge when alignment is confined to only the initial output tokens. Unfortunately, even with the introduction of deep safety alignment, determining the optimal safety depth remains an unresolved challenge. By leveraging the equivalence between autoregressive language models and Mark","authors_text":"Chia-Mu Yu, Ching-Chia Kao, Chun-Shien Lu, Chu-Song Chen","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-02T04:43:35Z","title":"Safety Alignment Depth in Large Language Models: A Markov Chain Perspective"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.00669","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":"sha256:8e79058e1473ecbe9302f6face0e1cb86cc4bcd0da5eea3b7035ecbccc2825fb","target":"record","created_at":"2026-07-05T10:08:36Z","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":"24b9eb61d65f3f86fc5314491b19dced431ace816d5511c197ca35dbf30cfa35","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-02T04:43:35Z","title_canon_sha256":"6f3f71568130e61e1c85a773b866fd5d352c0255830cb6edb6ce523388a01b87"},"schema_version":"1.0","source":{"id":"2502.00669","kind":"arxiv","version":1}},"canonical_sha256":"d0870694a8b19c000f95564cd448d84afea88d98e49a45c40129e3c3219dc5e3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d0870694a8b19c000f95564cd448d84afea88d98e49a45c40129e3c3219dc5e3","first_computed_at":"2026-07-05T10:08:36.889028Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:08:36.889028Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"nYe1mGtlQWQ/xZmbOdrvC1mL8z4WFB7ZomePlTEvtyb4gIcan+trOxojF6OYz7oMwUXHFq7kwKD2G6HoSEMBDw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:08:36.889455Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.00669","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8e79058e1473ecbe9302f6face0e1cb86cc4bcd0da5eea3b7035ecbccc2825fb","sha256:6ae4790ff691c4622623f625023f4df8245cea9a39cbbe68faba0f005477930d"],"state_sha256":"0dc64cb24671e2824a448fddf364062d69023b9a04a207edb43fb7c38aca6fac"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jAtGBjthd19owg428Dmda9q3l3zGjybsOamcUeSo0KOZfGHvjULZFKQNIgo34XL5Sr4PdLCCx7g1Fe6qnWjKBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T10:56:31.964191Z","bundle_sha256":"b6af48edc1edb4ba8571192be08545be07632abc654c96fea68994e442fbe495"}}