{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:VWQTV4OSHRINX7UMODELWKFR3G","short_pith_number":"pith:VWQTV4OS","canonical_record":{"source":{"id":"2412.03235","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-04T11:36:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e442c27e2af477da1254784dbbcf015ce9031ab04d45b9ea394eb0448548c09a","abstract_canon_sha256":"8adf742bbe01a94007d5e3a6fe82a2c1659f1e56a59d23f7ff09d3af8dbe1595"},"schema_version":"1.0"},"canonical_sha256":"ada13af1d23c50dbfe8c70c8bb28b1d9bc556fa0bd3811d38a51d06ad840aee3","source":{"kind":"arxiv","id":"2412.03235","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.03235","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"arxiv_version","alias_value":"2412.03235v2","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.03235","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"pith_short_12","alias_value":"VWQTV4OSHRIN","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"pith_short_16","alias_value":"VWQTV4OSHRINX7UM","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"pith_short_8","alias_value":"VWQTV4OS","created_at":"2026-07-05T10:38:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:VWQTV4OSHRINX7UMODELWKFR3G","target":"record","payload":{"canonical_record":{"source":{"id":"2412.03235","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-04T11:36:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e442c27e2af477da1254784dbbcf015ce9031ab04d45b9ea394eb0448548c09a","abstract_canon_sha256":"8adf742bbe01a94007d5e3a6fe82a2c1659f1e56a59d23f7ff09d3af8dbe1595"},"schema_version":"1.0"},"canonical_sha256":"ada13af1d23c50dbfe8c70c8bb28b1d9bc556fa0bd3811d38a51d06ad840aee3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:38:39.087982Z","signature_b64":"Hs5Yp3onJg1gydzfq0f0xOk54tlP1TleLUOKCCHjbVinC1gIlW5HivsqhfdevX6ab5LcKFLCCv9sMJ7NroaBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ada13af1d23c50dbfe8c70c8bb28b1d9bc556fa0bd3811d38a51d06ad840aee3","last_reissued_at":"2026-07-05T10:38:39.087513Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:38:39.087513Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.03235","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:38:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/O64hj5X+WLRMfIfKmmp2/dPswHNzfYXprPMsqQXv0KwVmDDhqwG5e7ASgCRQkPebZb8oO5o6klc/m2XuY1sBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T21:12:37.160945Z"},"content_sha256":"29299461a6158d84788215408dc5b4212ecedc66b54bdf76313afe9553b18f1e","schema_version":"1.0","event_id":"sha256:29299461a6158d84788215408dc5b4212ecedc66b54bdf76313afe9553b18f1e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:VWQTV4OSHRINX7UMODELWKFR3G","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Does Safety Training of LLMs Generalize to Semantically Related Natural Prompts?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Arun Suggala, Karthikeyan Shanmugam, Prateek Jain, Sravanti Addepalli, Yerram Varun","submitted_at":"2024-12-04T11:36:37Z","abstract_excerpt":"Large Language Models (LLMs) are known to be susceptible to crafted adversarial attacks or jailbreaks that lead to the generation of objectionable content despite being aligned to human preferences using safety fine-tuning methods. While the large dimensionality of input token space makes it inevitable to find adversarial prompts that can jailbreak these models, we aim to evaluate whether safety fine-tuned LLMs are safe against natural prompts which are semantically related to toxic seed prompts that elicit safe responses after alignment. We surprisingly find that popular aligned LLMs such as "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.03235","kind":"arxiv","version":2},"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/2412.03235/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":"2026-07-05T10:38:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wWbYxSbDEqgRORMlCmFCOz4yx7Jstp7B9Y+FMUTmOhrqgHXspY3h7Z2ZdNEA91DyKaeXifti6zLU8G3sFidbBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T21:12:37.161963Z"},"content_sha256":"55d38f1d8b8df1c28a051566a40a829abe292cb3d8b27732a326a5fd3bb7fcc1","schema_version":"1.0","event_id":"sha256:55d38f1d8b8df1c28a051566a40a829abe292cb3d8b27732a326a5fd3bb7fcc1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VWQTV4OSHRINX7UMODELWKFR3G/bundle.json","state_url":"https://pith.science/pith/VWQTV4OSHRINX7UMODELWKFR3G/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VWQTV4OSHRINX7UMODELWKFR3G/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":"2026-08-03T21:12:37Z","links":{"resolver":"https://pith.science/pith/VWQTV4OSHRINX7UMODELWKFR3G","bundle":"https://pith.science/pith/VWQTV4OSHRINX7UMODELWKFR3G/bundle.json","state":"https://pith.science/pith/VWQTV4OSHRINX7UMODELWKFR3G/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VWQTV4OSHRINX7UMODELWKFR3G/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:VWQTV4OSHRINX7UMODELWKFR3G","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":"8adf742bbe01a94007d5e3a6fe82a2c1659f1e56a59d23f7ff09d3af8dbe1595","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-04T11:36:37Z","title_canon_sha256":"e442c27e2af477da1254784dbbcf015ce9031ab04d45b9ea394eb0448548c09a"},"schema_version":"1.0","source":{"id":"2412.03235","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.03235","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"arxiv_version","alias_value":"2412.03235v2","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.03235","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"pith_short_12","alias_value":"VWQTV4OSHRIN","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"pith_short_16","alias_value":"VWQTV4OSHRINX7UM","created_at":"2026-07-05T10:38:39Z"},{"alias_kind":"pith_short_8","alias_value":"VWQTV4OS","created_at":"2026-07-05T10:38:39Z"}],"graph_snapshots":[{"event_id":"sha256:55d38f1d8b8df1c28a051566a40a829abe292cb3d8b27732a326a5fd3bb7fcc1","target":"graph","created_at":"2026-07-05T10:38:39Z","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/2412.03235/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are known to be susceptible to crafted adversarial attacks or jailbreaks that lead to the generation of objectionable content despite being aligned to human preferences using safety fine-tuning methods. While the large dimensionality of input token space makes it inevitable to find adversarial prompts that can jailbreak these models, we aim to evaluate whether safety fine-tuned LLMs are safe against natural prompts which are semantically related to toxic seed prompts that elicit safe responses after alignment. We surprisingly find that popular aligned LLMs such as ","authors_text":"Arun Suggala, Karthikeyan Shanmugam, Prateek Jain, Sravanti Addepalli, Yerram Varun","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-04T11:36:37Z","title":"Does Safety Training of LLMs Generalize to Semantically Related Natural Prompts?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.03235","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":"sha256:29299461a6158d84788215408dc5b4212ecedc66b54bdf76313afe9553b18f1e","target":"record","created_at":"2026-07-05T10:38:39Z","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":"8adf742bbe01a94007d5e3a6fe82a2c1659f1e56a59d23f7ff09d3af8dbe1595","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-04T11:36:37Z","title_canon_sha256":"e442c27e2af477da1254784dbbcf015ce9031ab04d45b9ea394eb0448548c09a"},"schema_version":"1.0","source":{"id":"2412.03235","kind":"arxiv","version":2}},"canonical_sha256":"ada13af1d23c50dbfe8c70c8bb28b1d9bc556fa0bd3811d38a51d06ad840aee3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ada13af1d23c50dbfe8c70c8bb28b1d9bc556fa0bd3811d38a51d06ad840aee3","first_computed_at":"2026-07-05T10:38:39.087513Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:38:39.087513Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Hs5Yp3onJg1gydzfq0f0xOk54tlP1TleLUOKCCHjbVinC1gIlW5HivsqhfdevX6ab5LcKFLCCv9sMJ7NroaBDw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:38:39.087982Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.03235","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:29299461a6158d84788215408dc5b4212ecedc66b54bdf76313afe9553b18f1e","sha256:55d38f1d8b8df1c28a051566a40a829abe292cb3d8b27732a326a5fd3bb7fcc1"],"state_sha256":"a31eff5706a38286e10f7356464a7930604a29a994ba4391232400ef4768d2a4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"m1AcabO1QDLzVydKZjmAbKbN+avvuuhiftyjngLMeQR9QKM3MWq6XOEFg4V8YuPDvR/SCnS5Rpwa0snXD3aMAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T21:12:37.169685Z","bundle_sha256":"6f02a757f1582aa7e09d0948d14093a1043b63b6bfe771e5e4ee92b140e961a9"}}