{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YLPYTJW4E7UTIOWV4SORV2VZNI","short_pith_number":"pith:YLPYTJW4","schema_version":"1.0","canonical_sha256":"c2df89a6dc27e9343ad5e49d1aeab96a2874daa1b9903edc3bfce00dafacb102","source":{"kind":"arxiv","id":"2503.01865","version":1},"attestation_state":"computed","paper":{"title":"Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Hongning Wang, Junxiao Yang, Minlie Huang, Shiyao Cui, Zhexin Zhang","submitted_at":"2025-02-25T07:47:41Z","abstract_excerpt":"Jailbreaking attacks can effectively induce unsafe behaviors in Large Language Models (LLMs); however, the transferability of these attacks across different models remains limited. This study aims to understand and enhance the transferability of gradient-based jailbreaking methods, which are among the standard approaches for attacking white-box models. Through a detailed analysis of the optimization process, we introduce a novel conceptual framework to elucidate transferability and identify superfluous constraints-specifically, the response pattern constraint and the token tail constraint-as s"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2503.01865","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-25T07:47:41Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"c0de976a2fc0b4ae56a5d9e11c79699496f6c018d39883ee6361162f6c45c12d","abstract_canon_sha256":"048157ef6568b09c4f01ce7ade1ca36e517c109057bfd7820f69c9027faa7457"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:23:31.470391Z","signature_b64":"NwfqcfjadgdDUa3B4w6uGw+xHX1ADOoSy54xYAwUqtc/hBpAqMPmZzjyvJRpS+NjVJvqA+Yh8Bx8XIdwR6WXAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c2df89a6dc27e9343ad5e49d1aeab96a2874daa1b9903edc3bfce00dafacb102","last_reissued_at":"2026-07-05T10:23:31.469736Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:23:31.469736Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Hongning Wang, Junxiao Yang, Minlie Huang, Shiyao Cui, Zhexin Zhang","submitted_at":"2025-02-25T07:47:41Z","abstract_excerpt":"Jailbreaking attacks can effectively induce unsafe behaviors in Large Language Models (LLMs); however, the transferability of these attacks across different models remains limited. This study aims to understand and enhance the transferability of gradient-based jailbreaking methods, which are among the standard approaches for attacking white-box models. Through a detailed analysis of the optimization process, we introduce a novel conceptual framework to elucidate transferability and identify superfluous constraints-specifically, the response pattern constraint and the token tail constraint-as s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.01865","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/2503.01865/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2503.01865","created_at":"2026-07-05T10:23:31.469811+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.01865v1","created_at":"2026-07-05T10:23:31.469811+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.01865","created_at":"2026-07-05T10:23:31.469811+00:00"},{"alias_kind":"pith_short_12","alias_value":"YLPYTJW4E7UT","created_at":"2026-07-05T10:23:31.469811+00:00"},{"alias_kind":"pith_short_16","alias_value":"YLPYTJW4E7UTIOWV","created_at":"2026-07-05T10:23:31.469811+00:00"},{"alias_kind":"pith_short_8","alias_value":"YLPYTJW4","created_at":"2026-07-05T10:23:31.469811+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.18543","citing_title":"SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models","ref_index":43,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI","json":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI.json","graph_json":"https://pith.science/api/pith-number/YLPYTJW4E7UTIOWV4SORV2VZNI/graph.json","events_json":"https://pith.science/api/pith-number/YLPYTJW4E7UTIOWV4SORV2VZNI/events.json","paper":"https://pith.science/paper/YLPYTJW4"},"agent_actions":{"view_html":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI","download_json":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI.json","view_paper":"https://pith.science/paper/YLPYTJW4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.01865&json=true","fetch_graph":"https://pith.science/api/pith-number/YLPYTJW4E7UTIOWV4SORV2VZNI/graph.json","fetch_events":"https://pith.science/api/pith-number/YLPYTJW4E7UTIOWV4SORV2VZNI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI/action/storage_attestation","attest_author":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI/action/author_attestation","sign_citation":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI/action/citation_signature","submit_replication":"https://pith.science/pith/YLPYTJW4E7UTIOWV4SORV2VZNI/action/replication_record"}},"created_at":"2026-07-05T10:23:31.469811+00:00","updated_at":"2026-07-05T10:23:31.469811+00:00"}