{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CDKYE3FTBAXVIDYNF76QKHDHOS","short_pith_number":"pith:CDKYE3FT","schema_version":"1.0","canonical_sha256":"10d5826cb3082f540f0d2ffd051c6774861ac0312ce8fff33548242532086c96","source":{"kind":"arxiv","id":"2502.19820","version":3},"attestation_state":"computed","paper":{"title":"Foot-In-The-Door: A Multi-turn Jailbreak for LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jinyuan Jia, Xiangyu Zhang, Xiaolong Jin, Zixuan Weng","submitted_at":"2025-02-27T06:49:16Z","abstract_excerpt":"Ensuring AI safety is crucial as large language models become increasingly integrated into real-world applications. A key challenge is jailbreak, where adversarial prompts bypass built-in safeguards to elicit harmful disallowed outputs. Inspired by psychological foot-in-the-door principles, we introduce FITD,a novel multi-turn jailbreak method that leverages the phenomenon where minor initial commitments lower resistance to more significant or more unethical transgressions. Our approach progressively escalates the malicious intent of user queries through intermediate bridge prompts and aligns "},"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":"2502.19820","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-27T06:49:16Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"42ee5fc19a4456cd7997e64d81c700f015bf003789cde43851b6ad7fdcc15d0f","abstract_canon_sha256":"43e25624b7195b2a1fed805e068fb64cdba9d567b286c3399482b83d778095ff"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:40:38.938904Z","signature_b64":"dMYW4jD56IHEML+DXSAnE9QKY3v7W+/7ttuFlr9ZascNWxaLYp+8b7hmhsKK5SQnOPE554NpnbGrbc6y1MWhAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"10d5826cb3082f540f0d2ffd051c6774861ac0312ce8fff33548242532086c96","last_reissued_at":"2026-07-05T10:40:38.938409Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:40:38.938409Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Foot-In-The-Door: A Multi-turn Jailbreak for LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jinyuan Jia, Xiangyu Zhang, Xiaolong Jin, Zixuan Weng","submitted_at":"2025-02-27T06:49:16Z","abstract_excerpt":"Ensuring AI safety is crucial as large language models become increasingly integrated into real-world applications. A key challenge is jailbreak, where adversarial prompts bypass built-in safeguards to elicit harmful disallowed outputs. Inspired by psychological foot-in-the-door principles, we introduce FITD,a novel multi-turn jailbreak method that leverages the phenomenon where minor initial commitments lower resistance to more significant or more unethical transgressions. Our approach progressively escalates the malicious intent of user queries through intermediate bridge prompts and aligns "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.19820","kind":"arxiv","version":3},"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/2502.19820/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":"2502.19820","created_at":"2026-07-05T10:40:38.938469+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.19820v3","created_at":"2026-07-05T10:40:38.938469+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.19820","created_at":"2026-07-05T10:40:38.938469+00:00"},{"alias_kind":"pith_short_12","alias_value":"CDKYE3FTBAXV","created_at":"2026-07-05T10:40:38.938469+00:00"},{"alias_kind":"pith_short_16","alias_value":"CDKYE3FTBAXVIDYN","created_at":"2026-07-05T10:40:38.938469+00:00"},{"alias_kind":"pith_short_8","alias_value":"CDKYE3FT","created_at":"2026-07-05T10:40:38.938469+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07808","citing_title":"Where Instruction Hierarchy Breaks: Diagnosing and Repairing Failures in Reasoning Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18988","citing_title":"Surviving the Unseen: Predictive Defense for Novel Multi-Turn Multimodal Attacks","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2509.10546","citing_title":"Learning to Conceal Risk: Controllable Multi-turn Red Teaming for LLMs in the Financial Domain","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22830","citing_title":"ChatInject: Abusing Chat Templates for Prompt Injection in LLM Agents","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10326","citing_title":"Jailbreaking the Matrix: Nullspace Steering for Controlled Model Subversion","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11309","citing_title":"The Salami Slicing Threat: Exploiting Cumulative Risks in LLM Systems","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15488","citing_title":"FineSteer: A Unified Framework for Fine-Grained Inference-Time Steering in Large Language Models","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02647","citing_title":"ContextualJailbreak: Evolutionary Red-Teaming via Simulated Conversational Priming","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS","json":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS.json","graph_json":"https://pith.science/api/pith-number/CDKYE3FTBAXVIDYNF76QKHDHOS/graph.json","events_json":"https://pith.science/api/pith-number/CDKYE3FTBAXVIDYNF76QKHDHOS/events.json","paper":"https://pith.science/paper/CDKYE3FT"},"agent_actions":{"view_html":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS","download_json":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS.json","view_paper":"https://pith.science/paper/CDKYE3FT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.19820&json=true","fetch_graph":"https://pith.science/api/pith-number/CDKYE3FTBAXVIDYNF76QKHDHOS/graph.json","fetch_events":"https://pith.science/api/pith-number/CDKYE3FTBAXVIDYNF76QKHDHOS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS/action/storage_attestation","attest_author":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS/action/author_attestation","sign_citation":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS/action/citation_signature","submit_replication":"https://pith.science/pith/CDKYE3FTBAXVIDYNF76QKHDHOS/action/replication_record"}},"created_at":"2026-07-05T10:40:38.938469+00:00","updated_at":"2026-07-05T10:40:38.938469+00:00"}