{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GNM5N2LABKQITLIIDWQ3O2LRCK","short_pith_number":"pith:GNM5N2LA","schema_version":"1.0","canonical_sha256":"3359d6e9600aa089ad081da1b7697112a517f60818cb3931efbe06cd65c784f4","source":{"kind":"arxiv","id":"2507.01831","version":1},"attestation_state":"computed","paper":{"title":"Out-of-Distribution Detection Methods Answer the Wrong Questions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrew Gordon Wilson, C. Bayan Bruss, Daohan Lu, Polina Kirichenko, Shikai Qiu, Tim G. J. Rudner, Yucen Lily Li","submitted_at":"2025-07-02T15:45:17Z","abstract_excerpt":"To detect distribution shifts and improve model safety, many out-of-distribution (OOD) detection methods rely on the predictive uncertainty or features of supervised models trained on in-distribution data. In this paper, we critically re-examine this popular family of OOD detection procedures, and we argue that these methods are fundamentally answering the wrong questions for OOD detection. There is no simple fix to this misalignment, since a classifier trained only on in-distribution classes cannot be expected to identify OOD points; for instance, a cat-dog classifier may confidently misclass"},"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":"2507.01831","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-07-02T15:45:17Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"7fdf5eb4103cb67f8455f8ae13d2c7a78661a40c28e1fa82600afe20bc859e55","abstract_canon_sha256":"43c3d5b94021198f68d12221eaf7f27dc16d13c3beb0de8294945c949fd862b7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:30:56.118514Z","signature_b64":"TIbThBQKMrv7SoNq5VQMamEdphZ3LECH8ApVZ/sSm+RtnsnBOyou4OoLyr7pbdo3yZSEPvs4PcJZjTL8WU09Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3359d6e9600aa089ad081da1b7697112a517f60818cb3931efbe06cd65c784f4","last_reissued_at":"2026-07-05T11:30:56.117712Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:30:56.117712Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Out-of-Distribution Detection Methods Answer the Wrong Questions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrew Gordon Wilson, C. Bayan Bruss, Daohan Lu, Polina Kirichenko, Shikai Qiu, Tim G. J. Rudner, Yucen Lily Li","submitted_at":"2025-07-02T15:45:17Z","abstract_excerpt":"To detect distribution shifts and improve model safety, many out-of-distribution (OOD) detection methods rely on the predictive uncertainty or features of supervised models trained on in-distribution data. In this paper, we critically re-examine this popular family of OOD detection procedures, and we argue that these methods are fundamentally answering the wrong questions for OOD detection. There is no simple fix to this misalignment, since a classifier trained only on in-distribution classes cannot be expected to identify OOD points; for instance, a cat-dog classifier may confidently misclass"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.01831","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/2507.01831/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":"2507.01831","created_at":"2026-07-05T11:30:56.117792+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.01831v1","created_at":"2026-07-05T11:30:56.117792+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.01831","created_at":"2026-07-05T11:30:56.117792+00:00"},{"alias_kind":"pith_short_12","alias_value":"GNM5N2LABKQI","created_at":"2026-07-05T11:30:56.117792+00:00"},{"alias_kind":"pith_short_16","alias_value":"GNM5N2LABKQITLII","created_at":"2026-07-05T11:30:56.117792+00:00"},{"alias_kind":"pith_short_8","alias_value":"GNM5N2LA","created_at":"2026-07-05T11:30:56.117792+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.17381","citing_title":"Beyond Binary Out-of-Distribution Detection: Characterizing Distributional Shifts with Multi-Statistic Diffusion Trajectories","ref_index":38,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK","json":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK.json","graph_json":"https://pith.science/api/pith-number/GNM5N2LABKQITLIIDWQ3O2LRCK/graph.json","events_json":"https://pith.science/api/pith-number/GNM5N2LABKQITLIIDWQ3O2LRCK/events.json","paper":"https://pith.science/paper/GNM5N2LA"},"agent_actions":{"view_html":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK","download_json":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK.json","view_paper":"https://pith.science/paper/GNM5N2LA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.01831&json=true","fetch_graph":"https://pith.science/api/pith-number/GNM5N2LABKQITLIIDWQ3O2LRCK/graph.json","fetch_events":"https://pith.science/api/pith-number/GNM5N2LABKQITLIIDWQ3O2LRCK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK/action/storage_attestation","attest_author":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK/action/author_attestation","sign_citation":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK/action/citation_signature","submit_replication":"https://pith.science/pith/GNM5N2LABKQITLIIDWQ3O2LRCK/action/replication_record"}},"created_at":"2026-07-05T11:30:56.117792+00:00","updated_at":"2026-07-05T11:30:56.117792+00:00"}