{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:QBBVIGVVWEZTRRY4PITJKK765S","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":"8f71cfed80622b3dc46ddf216ce7c1c45d9071661ef6da5bcd999c5933aae4a5","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2023-12-21T21:22:41Z","title_canon_sha256":"a43ded27f1541836e9eb70b29308870d2532072aa0c09a7a0770ad01afd66b6a"},"schema_version":"1.0","source":{"id":"2312.14302","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.14302","created_at":"2026-07-05T08:51:45Z"},{"alias_kind":"arxiv_version","alias_value":"2312.14302v2","created_at":"2026-07-05T08:51:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.14302","created_at":"2026-07-05T08:51:45Z"},{"alias_kind":"pith_short_12","alias_value":"QBBVIGVVWEZT","created_at":"2026-07-05T08:51:45Z"},{"alias_kind":"pith_short_16","alias_value":"QBBVIGVVWEZTRRY4","created_at":"2026-07-05T08:51:45Z"},{"alias_kind":"pith_short_8","alias_value":"QBBVIGVV","created_at":"2026-07-05T08:51:45Z"}],"graph_snapshots":[{"event_id":"sha256:d7c45eccd02c3e4ba6fc0324b6e84bed65a11242b8503bb7dd441365ba5cc5fd","target":"graph","created_at":"2026-07-05T08:51:45Z","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/2312.14302/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language model attacks typically assume one of two extreme threat models: full white-box access to model weights, or black-box access limited to a text generation API. However, real-world APIs are often more flexible than just text generation: these APIs expose \"gray-box\" access leading to new threat vectors. To explore this, we red-team three new functionalities exposed in the GPT-4 APIs: fine-tuning, function calling and knowledge retrieval. We find that fine-tuning a model on as few as 15 harmful examples or 100 benign examples can remove core safeguards from GPT-4, enabling a range of harm","authors_text":"Adam Gleave, Euan McLean, Kellin Pelrine, Micha{\\l} Zaj\\k{a}c, Mohammad Taufeeque","cross_cats":["cs.AI","cs.CL","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2023-12-21T21:22:41Z","title":"Exploiting Novel GPT-4 APIs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.14302","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":"sha256:40a9ff2ae98a17499e0f2531c29f13d98e10586363d8a17cec07e41db4729fc4","target":"record","created_at":"2026-07-05T08:51:45Z","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":"8f71cfed80622b3dc46ddf216ce7c1c45d9071661ef6da5bcd999c5933aae4a5","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2023-12-21T21:22:41Z","title_canon_sha256":"a43ded27f1541836e9eb70b29308870d2532072aa0c09a7a0770ad01afd66b6a"},"schema_version":"1.0","source":{"id":"2312.14302","kind":"arxiv","version":2}},"canonical_sha256":"8043541ab5b13338c71c7a26952bfeec949637184746909fb8eebac47ec68ea6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8043541ab5b13338c71c7a26952bfeec949637184746909fb8eebac47ec68ea6","first_computed_at":"2026-07-05T08:51:45.468939Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:51:45.468939Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"TuyEbzddJVOyzvrK7ZvdqZv2MTGR66JVH/zpAGngdLTMDNSxcOLOeLEdCODXxfp8PsX6n2VSGRdCPZ2yQ7+wDA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:51:45.469409Z","signed_message":"canonical_sha256_bytes"},"source_id":"2312.14302","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:40a9ff2ae98a17499e0f2531c29f13d98e10586363d8a17cec07e41db4729fc4","sha256:d7c45eccd02c3e4ba6fc0324b6e84bed65a11242b8503bb7dd441365ba5cc5fd"],"state_sha256":"ea24cc6744085dafafa857592617a22ca509ff9b667eca62bbe10a987e194a2a"}