{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NTFZTWS3SDBLNPNJ3FZGUCXZN3","short_pith_number":"pith:NTFZTWS3","schema_version":"1.0","canonical_sha256":"6ccb99da5b90c2b6bda9d9726a0af96ef74a4c88b756d8c15896c74fabeecb89","source":{"kind":"arxiv","id":"2402.03303","version":1},"attestation_state":"computed","paper":{"title":"Nevermind: Instruction Override and Moderation in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Edward Kim","submitted_at":"2024-02-05T18:58:19Z","abstract_excerpt":"Given the impressive capabilities of recent Large Language Models (LLMs), we investigate and benchmark the most popular proprietary and different sized open source models on the task of explicit instruction following in conflicting situations, e.g. overrides. These include the ability of the model to override the knowledge within the weights of the model, the ability to override (or moderate) extracted knowledge in the prompt, and lastly the ability to perform a full jailbreak. Experimentation performed suggest several key findings to improve instruction following - larger models perform the b"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2402.03303","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-05T18:58:19Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"be9c96f03d4d1b851a05714d8a297a0457cff67dcb97e8d880168a065e17cbf4","abstract_canon_sha256":"7c3873de23c647fd25b31c3e9e6b8ae852ac316bc03ae06cc5a5904b05cb67f9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:41:33.688726Z","signature_b64":"x897oHFeAvzD8igzbOtm+y6zhyCzVjoevUFECI/wSta34HFYaQYy/2V+PQ31deambvOxwdso/qV/YG4FzTGFAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6ccb99da5b90c2b6bda9d9726a0af96ef74a4c88b756d8c15896c74fabeecb89","last_reissued_at":"2026-07-05T07:41:33.688145Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:41:33.688145Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Nevermind: Instruction Override and Moderation in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Edward Kim","submitted_at":"2024-02-05T18:58:19Z","abstract_excerpt":"Given the impressive capabilities of recent Large Language Models (LLMs), we investigate and benchmark the most popular proprietary and different sized open source models on the task of explicit instruction following in conflicting situations, e.g. overrides. These include the ability of the model to override the knowledge within the weights of the model, the ability to override (or moderate) extracted knowledge in the prompt, and lastly the ability to perform a full jailbreak. Experimentation performed suggest several key findings to improve instruction following - larger models perform the b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.03303","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2402.03303/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2402.03303","created_at":"2026-07-05T07:41:33.688207+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.03303v1","created_at":"2026-07-05T07:41:33.688207+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.03303","created_at":"2026-07-05T07:41:33.688207+00:00"},{"alias_kind":"pith_short_12","alias_value":"NTFZTWS3SDBL","created_at":"2026-07-05T07:41:33.688207+00:00"},{"alias_kind":"pith_short_16","alias_value":"NTFZTWS3SDBLNPNJ","created_at":"2026-07-05T07:41:33.688207+00:00"},{"alias_kind":"pith_short_8","alias_value":"NTFZTWS3","created_at":"2026-07-05T07:41:33.688207+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.24428","citing_title":"Model Unlearning via Sparse Autoencoder Subspace Guided Projections","ref_index":10,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3","json":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3.json","graph_json":"https://pith.science/api/pith-number/NTFZTWS3SDBLNPNJ3FZGUCXZN3/graph.json","events_json":"https://pith.science/api/pith-number/NTFZTWS3SDBLNPNJ3FZGUCXZN3/events.json","paper":"https://pith.science/paper/NTFZTWS3"},"agent_actions":{"view_html":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3","download_json":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3.json","view_paper":"https://pith.science/paper/NTFZTWS3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.03303&json=true","fetch_graph":"https://pith.science/api/pith-number/NTFZTWS3SDBLNPNJ3FZGUCXZN3/graph.json","fetch_events":"https://pith.science/api/pith-number/NTFZTWS3SDBLNPNJ3FZGUCXZN3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3/action/storage_attestation","attest_author":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3/action/author_attestation","sign_citation":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3/action/citation_signature","submit_replication":"https://pith.science/pith/NTFZTWS3SDBLNPNJ3FZGUCXZN3/action/replication_record"}},"created_at":"2026-07-05T07:41:33.688207+00:00","updated_at":"2026-07-05T07:41:33.688207+00:00"}