{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:JADIFN7LT35UP4BSEQM7ORMVPL","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":"fae8eec9459dcbd1012d47236406998427e264e5f09a050fb2bade880e0db096","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-04-14T20:30:41Z","title_canon_sha256":"62bfcc1f59afcb32c3c3d271c6dbdb5a886ae9a4102cbb4eef36a2ed72e1ba60"},"schema_version":"1.0","source":{"id":"2504.10694","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.10694","created_at":"2026-07-05T10:49:22Z"},{"alias_kind":"arxiv_version","alias_value":"2504.10694v1","created_at":"2026-07-05T10:49:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.10694","created_at":"2026-07-05T10:49:22Z"},{"alias_kind":"pith_short_12","alias_value":"JADIFN7LT35U","created_at":"2026-07-05T10:49:22Z"},{"alias_kind":"pith_short_16","alias_value":"JADIFN7LT35UP4BS","created_at":"2026-07-05T10:49:22Z"},{"alias_kind":"pith_short_8","alias_value":"JADIFN7L","created_at":"2026-07-05T10:49:22Z"}],"graph_snapshots":[{"event_id":"sha256:cbf0a5dab79aeb7f287a8004be92ffaf404cc6bb997928d7404ac8c6deda23a0","target":"graph","created_at":"2026-07-05T10:49:22Z","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/2504.10694/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Jailbreak attacks bypass the guardrails of large language models to produce harmful outputs. In this paper, we ask whether the model outputs produced by existing jailbreaks are actually useful. For example, when jailbreaking a model to give instructions for building a bomb, does the jailbreak yield good instructions? Since the utility of most unsafe answers (e.g., bomb instructions) is hard to evaluate rigorously, we build new jailbreak evaluation sets with known ground truth answers, by aligning models to refuse questions related to benign and easy-to-evaluate topics (e.g., biology or math). ","authors_text":"Florian Tram\\`er, Jie Zhang, Kristina Nikoli\\'c, Luze Sun","cross_cats":["cs.AI","cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-04-14T20:30:41Z","title":"The Jailbreak Tax: How Useful are Your Jailbreak Outputs?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.10694","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:fcf7a013b914f52787d9d27119b857e173569c54bf26fe09188c950fcc3cceb7","target":"record","created_at":"2026-07-05T10:49:22Z","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":"fae8eec9459dcbd1012d47236406998427e264e5f09a050fb2bade880e0db096","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-04-14T20:30:41Z","title_canon_sha256":"62bfcc1f59afcb32c3c3d271c6dbdb5a886ae9a4102cbb4eef36a2ed72e1ba60"},"schema_version":"1.0","source":{"id":"2504.10694","kind":"arxiv","version":1}},"canonical_sha256":"480682b7eb9efb47f0322419f745957ad5e177d85cea11b0f6bc962d1811ffcd","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"480682b7eb9efb47f0322419f745957ad5e177d85cea11b0f6bc962d1811ffcd","first_computed_at":"2026-07-05T10:49:22.478663Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:49:22.478663Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Ie48Z53kB56/sT3dci2+t3fyAgpQljJ6WURWuTAQvIfkmY6xtH+u8vZipF3PDV85OelISOqnh6ysCKPthOrkCA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:49:22.479160Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.10694","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:fcf7a013b914f52787d9d27119b857e173569c54bf26fe09188c950fcc3cceb7","sha256:cbf0a5dab79aeb7f287a8004be92ffaf404cc6bb997928d7404ac8c6deda23a0"],"state_sha256":"e35ec8fc266518657b627da09a9a73a4e694e63a37eec39640685a50a174564f"}