{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:SYSRC46WM6FPEBFZOYGWBQ53SB","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":"d6e3f608d67c8c5c7c500a443b232c011b43222bfa842361eeeb5018195ed006","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-16T15:03:33Z","title_canon_sha256":"e71f8a572654e88e3f69edc3fe04b0300bc203354ee7016735a6b2b5254872d8"},"schema_version":"1.0","source":{"id":"2508.12072","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.12072","created_at":"2026-07-05T11:57:58Z"},{"alias_kind":"arxiv_version","alias_value":"2508.12072v2","created_at":"2026-07-05T11:57:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.12072","created_at":"2026-07-05T11:57:58Z"},{"alias_kind":"pith_short_12","alias_value":"SYSRC46WM6FP","created_at":"2026-07-05T11:57:58Z"},{"alias_kind":"pith_short_16","alias_value":"SYSRC46WM6FPEBFZ","created_at":"2026-07-05T11:57:58Z"},{"alias_kind":"pith_short_8","alias_value":"SYSRC46W","created_at":"2026-07-05T11:57:58Z"}],"graph_snapshots":[{"event_id":"sha256:e59a73b6064b8b58de39e97b0b091ca166f64d88bd58a9d19c33c3689cc0de48","target":"graph","created_at":"2026-07-05T11:57:58Z","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/2508.12072/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite extensive safety-tuning, large language models (LLMs) remain vulnerable to jailbreak attacks via adversarially crafted instructions, reflecting a persistent trade-off between safety and task performance. In this work, we propose Intent-FT, a simple and lightweight fine-tuning approach that explicitly trains LLMs to infer the underlying intent of an instruction before responding. By fine-tuning on a targeted set of adversarial instructions, Intent-FT enables LLMs to generalize intent deduction to unseen attacks, thereby substantially improving their robustness. We comprehensively evalua","authors_text":"Erik Cambria, Ranjan Satapathy, Wei Jie Yeo","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-16T15:03:33Z","title":"Mitigating Jailbreaks with Intent-Aware LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.12072","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":"sha256:2bbf98c1331b76eea1b1fadd2d992032995ad65230787ba9221e96fba00ec25d","target":"record","created_at":"2026-07-05T11:57:58Z","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":"d6e3f608d67c8c5c7c500a443b232c011b43222bfa842361eeeb5018195ed006","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-16T15:03:33Z","title_canon_sha256":"e71f8a572654e88e3f69edc3fe04b0300bc203354ee7016735a6b2b5254872d8"},"schema_version":"1.0","source":{"id":"2508.12072","kind":"arxiv","version":2}},"canonical_sha256":"96251173d6678af204b9760d60c3bb90565a31d4f434ebde474c03b2cb04b5e4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"96251173d6678af204b9760d60c3bb90565a31d4f434ebde474c03b2cb04b5e4","first_computed_at":"2026-07-05T11:57:58.939024Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:57:58.939024Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"x+sGmPnBPO0pA/Fzs9xIUD9bFUqLbLEX79mQh+9xTbbkaulzioQUaYpTuCmbwfuHaBsm9KV6oKP7v25MGwDVAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:57:58.939516Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.12072","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2bbf98c1331b76eea1b1fadd2d992032995ad65230787ba9221e96fba00ec25d","sha256:e59a73b6064b8b58de39e97b0b091ca166f64d88bd58a9d19c33c3689cc0de48"],"state_sha256":"3e953b7b1555a25f7935ea8ed645ee0438f1d659a519f55e88ae076c17bb9fb8"}