{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:JDDSNZN4VANRXQT4PU3EU3COE4","short_pith_number":"pith:JDDSNZN4","canonical_record":{"source":{"id":"2406.06622","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-07T15:37:15Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"a91ec1345817c73b9210646a2648fd89bc37ed84004cf4a527ce99ceabaac126","abstract_canon_sha256":"0c95526c7b7d700de31afd161787bf210ce7588079655827b5897606b95980a2"},"schema_version":"1.0"},"canonical_sha256":"48c726e5bca81b1bc27c7d364a6c4e271069c5e9c3ec0b779ab405d6266ee233","source":{"kind":"arxiv","id":"2406.06622","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.06622","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"arxiv_version","alias_value":"2406.06622v1","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.06622","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"pith_short_12","alias_value":"JDDSNZN4VANR","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"pith_short_16","alias_value":"JDDSNZN4VANRXQT4","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"pith_short_8","alias_value":"JDDSNZN4","created_at":"2026-07-05T08:29:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:JDDSNZN4VANRXQT4PU3EU3COE4","target":"record","payload":{"canonical_record":{"source":{"id":"2406.06622","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-07T15:37:15Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"a91ec1345817c73b9210646a2648fd89bc37ed84004cf4a527ce99ceabaac126","abstract_canon_sha256":"0c95526c7b7d700de31afd161787bf210ce7588079655827b5897606b95980a2"},"schema_version":"1.0"},"canonical_sha256":"48c726e5bca81b1bc27c7d364a6c4e271069c5e9c3ec0b779ab405d6266ee233","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:29:48.222891Z","signature_b64":"jbi1bVQRP9Q9sb9mUfPh0lP76NzzxuQ5jXXhv6kyF0fR62pFtMtRlkQ3HYsvjuYE/BO3PGa4zFbnV1yEqFmNBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"48c726e5bca81b1bc27c7d364a6c4e271069c5e9c3ec0b779ab405d6266ee233","last_reissued_at":"2026-07-05T08:29:48.222533Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:29:48.222533Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.06622","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":"2026-07-05T08:29:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nghmHcXjvhBgEVK3VfFAS/sX5kEKAJiNYSXt3SQ0H5LZVCQyImiHwccAXChe0N1RKhpU+I8q2lQcVtmHP2KVAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T19:18:42.617520Z"},"content_sha256":"6b23a9588b4f6959bbf4b91fe7689db8665763dfb2abe4e0c0abf8fda64c524d","schema_version":"1.0","event_id":"sha256:6b23a9588b4f6959bbf4b91fe7689db8665763dfb2abe4e0c0abf8fda64c524d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:JDDSNZN4VANRXQT4PU3EU3COE4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Adversarial Tuning: Defending Against Jailbreak Attacks for LLMs","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.CL","authors_text":"Fan Liu, Hao Liu, Zhao Xu","submitted_at":"2024-06-07T15:37:15Z","abstract_excerpt":"Although safely enhanced Large Language Models (LLMs) have achieved remarkable success in tackling various complex tasks in a zero-shot manner, they remain susceptible to jailbreak attacks, particularly the unknown jailbreak attack. To enhance LLMs' generalized defense capabilities, we propose a two-stage adversarial tuning framework, which generates adversarial prompts to explore worst-case scenarios by optimizing datasets containing pairs of adversarial prompts and their safe responses. In the first stage, we introduce the hierarchical meta-universal adversarial prompt learning to efficientl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.06622","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/2406.06622/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":"2026-07-05T08:29:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CumObxvA+YOmmT1eruluAc3cZWCbmIcCMywQeVqSKcf3R47TduSE6FSxxVCV7gkTMvyDjbp3qJftNgxuxF5xBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T19:18:42.618509Z"},"content_sha256":"9034055a6b83f63204f3deffad44bfb5e870be92a0f651f65f0d9e96d08a4ebe","schema_version":"1.0","event_id":"sha256:9034055a6b83f63204f3deffad44bfb5e870be92a0f651f65f0d9e96d08a4ebe"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JDDSNZN4VANRXQT4PU3EU3COE4/bundle.json","state_url":"https://pith.science/pith/JDDSNZN4VANRXQT4PU3EU3COE4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JDDSNZN4VANRXQT4PU3EU3COE4/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":"2026-08-13T19:18:42Z","links":{"resolver":"https://pith.science/pith/JDDSNZN4VANRXQT4PU3EU3COE4","bundle":"https://pith.science/pith/JDDSNZN4VANRXQT4PU3EU3COE4/bundle.json","state":"https://pith.science/pith/JDDSNZN4VANRXQT4PU3EU3COE4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JDDSNZN4VANRXQT4PU3EU3COE4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:JDDSNZN4VANRXQT4PU3EU3COE4","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":"0c95526c7b7d700de31afd161787bf210ce7588079655827b5897606b95980a2","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-07T15:37:15Z","title_canon_sha256":"a91ec1345817c73b9210646a2648fd89bc37ed84004cf4a527ce99ceabaac126"},"schema_version":"1.0","source":{"id":"2406.06622","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.06622","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"arxiv_version","alias_value":"2406.06622v1","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.06622","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"pith_short_12","alias_value":"JDDSNZN4VANR","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"pith_short_16","alias_value":"JDDSNZN4VANRXQT4","created_at":"2026-07-05T08:29:48Z"},{"alias_kind":"pith_short_8","alias_value":"JDDSNZN4","created_at":"2026-07-05T08:29:48Z"}],"graph_snapshots":[{"event_id":"sha256:9034055a6b83f63204f3deffad44bfb5e870be92a0f651f65f0d9e96d08a4ebe","target":"graph","created_at":"2026-07-05T08:29:48Z","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/2406.06622/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Although safely enhanced Large Language Models (LLMs) have achieved remarkable success in tackling various complex tasks in a zero-shot manner, they remain susceptible to jailbreak attacks, particularly the unknown jailbreak attack. To enhance LLMs' generalized defense capabilities, we propose a two-stage adversarial tuning framework, which generates adversarial prompts to explore worst-case scenarios by optimizing datasets containing pairs of adversarial prompts and their safe responses. In the first stage, we introduce the hierarchical meta-universal adversarial prompt learning to efficientl","authors_text":"Fan Liu, Hao Liu, Zhao Xu","cross_cats":["cs.AI","cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-07T15:37:15Z","title":"Adversarial Tuning: Defending Against Jailbreak Attacks for LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.06622","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":"sha256:6b23a9588b4f6959bbf4b91fe7689db8665763dfb2abe4e0c0abf8fda64c524d","target":"record","created_at":"2026-07-05T08:29:48Z","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":"0c95526c7b7d700de31afd161787bf210ce7588079655827b5897606b95980a2","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-07T15:37:15Z","title_canon_sha256":"a91ec1345817c73b9210646a2648fd89bc37ed84004cf4a527ce99ceabaac126"},"schema_version":"1.0","source":{"id":"2406.06622","kind":"arxiv","version":1}},"canonical_sha256":"48c726e5bca81b1bc27c7d364a6c4e271069c5e9c3ec0b779ab405d6266ee233","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"48c726e5bca81b1bc27c7d364a6c4e271069c5e9c3ec0b779ab405d6266ee233","first_computed_at":"2026-07-05T08:29:48.222533Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:29:48.222533Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jbi1bVQRP9Q9sb9mUfPh0lP76NzzxuQ5jXXhv6kyF0fR62pFtMtRlkQ3HYsvjuYE/BO3PGa4zFbnV1yEqFmNBg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:29:48.222891Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.06622","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6b23a9588b4f6959bbf4b91fe7689db8665763dfb2abe4e0c0abf8fda64c524d","sha256:9034055a6b83f63204f3deffad44bfb5e870be92a0f651f65f0d9e96d08a4ebe"],"state_sha256":"3070928526162de1b6464fc3f29e743bca537e545915a6383dece63d2c231e27"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dm8GBGnWZPaFvlXlKk3Nh3PRictgXmktvRPuNHQPvNaSRtiHshXbv4pVxdJ9UUvMxJQ8OEh/VREDYppVZ/BaDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-13T19:18:42.696600Z","bundle_sha256":"126bf5ee8ad531bd8f35a1914d49459cb904d394b6f3fc92ec5a66e651b1d322"}}