{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:2IRUGJIMQFFHEVKS4XO67O3JVA","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":"e944ce90ed7a48372db694708a48e6d58d9397cbded33507e698b45173a41cdc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-02T07:40:56Z","title_canon_sha256":"2881475bc8f58c043538b7e2099820d4f5bd030c54e9fc4fff7ef93adbda2c66"},"schema_version":"1.0","source":{"id":"2410.01294","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.01294","created_at":"2026-07-05T11:01:01Z"},{"alias_kind":"arxiv_version","alias_value":"2410.01294v3","created_at":"2026-07-05T11:01:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.01294","created_at":"2026-07-05T11:01:01Z"},{"alias_kind":"pith_short_12","alias_value":"2IRUGJIMQFFH","created_at":"2026-07-05T11:01:01Z"},{"alias_kind":"pith_short_16","alias_value":"2IRUGJIMQFFHEVKS","created_at":"2026-07-05T11:01:01Z"},{"alias_kind":"pith_short_8","alias_value":"2IRUGJIM","created_at":"2026-07-05T11:01:01Z"}],"graph_snapshots":[{"event_id":"sha256:a20d58d4cc2c6f6a7a5494fd30614c1c91fb4f77ad8802ef13bf6fff5e1c5e57","target":"graph","created_at":"2026-07-05T11:01:01Z","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/2410.01294/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite extensive safety measures, LLMs are vulnerable to adversarial inputs, or jailbreaks, which can elicit unsafe behaviors. In this work, we introduce bijection learning, a powerful attack algorithm which automatically fuzzes LLMs for safety vulnerabilities using randomly-generated encodings whose complexity can be tightly controlled. We leverage in-context learning to teach models bijective encodings, pass encoded queries to the model to bypass built-in safety mechanisms, and finally decode responses back into English. Our attack is extremely effective on a wide range of frontier language","authors_text":"Brian R.Y. Huang, Leonard Tang, Maximilian Li","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-02T07:40:56Z","title":"Endless Jailbreaks with Bijection Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.01294","kind":"arxiv","version":3},"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":"sha256:302ba3a6c414f8cc1a6d7107790f089edca6efadd9165aab312157983944d249","target":"record","created_at":"2026-07-05T11:01:01Z","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":"e944ce90ed7a48372db694708a48e6d58d9397cbded33507e698b45173a41cdc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-02T07:40:56Z","title_canon_sha256":"2881475bc8f58c043538b7e2099820d4f5bd030c54e9fc4fff7ef93adbda2c66"},"schema_version":"1.0","source":{"id":"2410.01294","kind":"arxiv","version":3}},"canonical_sha256":"d22343250c814a725552e5ddefbb69a80dc6b909167ac6c2339bd1c821d3217f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d22343250c814a725552e5ddefbb69a80dc6b909167ac6c2339bd1c821d3217f","first_computed_at":"2026-07-05T11:01:01.846974Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:01:01.846974Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"uUOo9gPnL72yQPLUzjqdYSrqBzrO9Oe8wTzzfLpy/PycCwgdx28scSEuAfA1VPR9WVSRKOijqKr+9cL1EzWeCA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:01:01.847445Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.01294","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:302ba3a6c414f8cc1a6d7107790f089edca6efadd9165aab312157983944d249","sha256:a20d58d4cc2c6f6a7a5494fd30614c1c91fb4f77ad8802ef13bf6fff5e1c5e57"],"state_sha256":"2d58dddc6a90478ef72355a2971a125a6cf2f4ada4902b6cd68eb45415bf8506"}