{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:KP467U67NEJXI4SXSCSPQ5KUEE","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":"f46879deef3cbf79e062e18921a1468388adf2f7d4d807314bc2ba582fe432a7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-31T19:45:15Z","title_canon_sha256":"340ba3ceb079cdf76d68b6be9897bcafa037ad856579a1a7763a1cf2a12539a5"},"schema_version":"1.0","source":{"id":"2311.00117","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.00117","created_at":"2026-07-05T08:23:57Z"},{"alias_kind":"arxiv_version","alias_value":"2311.00117v3","created_at":"2026-07-05T08:23:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.00117","created_at":"2026-07-05T08:23:57Z"},{"alias_kind":"pith_short_12","alias_value":"KP467U67NEJX","created_at":"2026-07-05T08:23:57Z"},{"alias_kind":"pith_short_16","alias_value":"KP467U67NEJXI4SX","created_at":"2026-07-05T08:23:57Z"},{"alias_kind":"pith_short_8","alias_value":"KP467U67","created_at":"2026-07-05T08:23:57Z"}],"graph_snapshots":[{"event_id":"sha256:4314c1e3dec486d169cc6f9d1283d2d4fc9d4f22594236fabc5418fd45aad47f","target":"graph","created_at":"2026-07-05T08:23:57Z","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/2311.00117/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Llama 2-Chat is a collection of large language models that Meta developed and released to the public. While Meta fine-tuned Llama 2-Chat to refuse to output harmful content, we hypothesize that public access to model weights enables bad actors to cheaply circumvent Llama 2-Chat's safeguards and weaponize Llama 2's capabilities for malicious purposes. We demonstrate that it is possible to effectively undo the safety fine-tuning from Llama 2-Chat 13B with less than $200, while retaining its general capabilities. Our results demonstrate that safety-fine tuning is ineffective at preventing misuse ","authors_text":"Charlie Rogers-Smith, Jeffrey Ladish, Pranav Gade, Simon Lermen","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-31T19:45:15Z","title":"BadLlama: cheaply removing safety fine-tuning from Llama 2-Chat 13B"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.00117","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:ac94c6ebb0a31b3abc21348b03b4350a105a0c9a4503ef12286b86109b6205d6","target":"record","created_at":"2026-07-05T08:23:57Z","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":"f46879deef3cbf79e062e18921a1468388adf2f7d4d807314bc2ba582fe432a7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-31T19:45:15Z","title_canon_sha256":"340ba3ceb079cdf76d68b6be9897bcafa037ad856579a1a7763a1cf2a12539a5"},"schema_version":"1.0","source":{"id":"2311.00117","kind":"arxiv","version":3}},"canonical_sha256":"53f9efd3df691374725790a4f875542131f38a4adf12a0a05fabd68c966e9558","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"53f9efd3df691374725790a4f875542131f38a4adf12a0a05fabd68c966e9558","first_computed_at":"2026-07-05T08:23:57.262447Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:23:57.262447Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"buHa4B9FshVTvFS7SXj+5SSCPHMeYEL0asWGvCuixzcVAlARflf0Bq0ZEsU15dziWpdK0lsB38Pv0NHkDhiKDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:23:57.262870Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.00117","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ac94c6ebb0a31b3abc21348b03b4350a105a0c9a4503ef12286b86109b6205d6","sha256:4314c1e3dec486d169cc6f9d1283d2d4fc9d4f22594236fabc5418fd45aad47f"],"state_sha256":"5918b959b75ac59e529157b045cdf686265d795bc5bc125a3970cb286f9c1ff6"}