{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:CHWLRHWPLSIHGYJBWS7CGBQJ3U","short_pith_number":"pith:CHWLRHWP","schema_version":"1.0","canonical_sha256":"11ecb89ecf5c90736121b4be230609dd090c8aed29d67352f8f81972f073afdb","source":{"kind":"arxiv","id":"2403.01485","version":2},"attestation_state":"computed","paper":{"title":"Approximations to the Fisher Information Metric of Deep Generative Models for Out-Of-Distribution Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"stat.ML","authors_text":"Chris Holmes, Christopher Williams, Fabian Falck, Sam Dauncey","submitted_at":"2024-03-03T11:36:35Z","abstract_excerpt":"Likelihood-based deep generative models such as score-based diffusion models and variational autoencoders are state-of-the-art machine learning models approximating high-dimensional distributions of data such as images, text, or audio. One of many downstream tasks they can be naturally applied to is out-of-distribution (OOD) detection. However, seminal work by Nalisnick et al. which we reproduce showed that deep generative models consistently infer higher log-likelihoods for OOD data than data they were trained on, marking an open problem. In this work, we analyse using the gradient of a data "},"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":"2403.01485","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-03-03T11:36:35Z","cross_cats_sorted":["cs.CV","cs.LG"],"title_canon_sha256":"8ae85b6fd4fbfdfc90d4e6b50f464900aaeaa35ff3f36a12cbbc3b187cb8bd20","abstract_canon_sha256":"3236c352399033eb484189cf1eddd7ead0a18a1d32e548b971e5b7366f626728"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:23:19.368127Z","signature_b64":"Kl/eChAorDLTfSe5HWp3uD5cBhhV+29pwz0bDuiRrjK56KRUNNT6MgWq72F+2gn++snUdCyN97I6lTh1wrEfBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"11ecb89ecf5c90736121b4be230609dd090c8aed29d67352f8f81972f073afdb","last_reissued_at":"2026-07-05T08:23:19.367665Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:23:19.367665Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Approximations to the Fisher Information Metric of Deep Generative Models for Out-Of-Distribution Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"stat.ML","authors_text":"Chris Holmes, Christopher Williams, Fabian Falck, Sam Dauncey","submitted_at":"2024-03-03T11:36:35Z","abstract_excerpt":"Likelihood-based deep generative models such as score-based diffusion models and variational autoencoders are state-of-the-art machine learning models approximating high-dimensional distributions of data such as images, text, or audio. One of many downstream tasks they can be naturally applied to is out-of-distribution (OOD) detection. However, seminal work by Nalisnick et al. which we reproduce showed that deep generative models consistently infer higher log-likelihoods for OOD data than data they were trained on, marking an open problem. In this work, we analyse using the gradient of a data "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.01485","kind":"arxiv","version":2},"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/2403.01485/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":"2403.01485","created_at":"2026-07-05T08:23:19.367722+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.01485v2","created_at":"2026-07-05T08:23:19.367722+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.01485","created_at":"2026-07-05T08:23:19.367722+00:00"},{"alias_kind":"pith_short_12","alias_value":"CHWLRHWPLSIH","created_at":"2026-07-05T08:23:19.367722+00:00"},{"alias_kind":"pith_short_16","alias_value":"CHWLRHWPLSIHGYJB","created_at":"2026-07-05T08:23:19.367722+00:00"},{"alias_kind":"pith_short_8","alias_value":"CHWLRHWP","created_at":"2026-07-05T08:23:19.367722+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":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U","json":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U.json","graph_json":"https://pith.science/api/pith-number/CHWLRHWPLSIHGYJBWS7CGBQJ3U/graph.json","events_json":"https://pith.science/api/pith-number/CHWLRHWPLSIHGYJBWS7CGBQJ3U/events.json","paper":"https://pith.science/paper/CHWLRHWP"},"agent_actions":{"view_html":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U","download_json":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U.json","view_paper":"https://pith.science/paper/CHWLRHWP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.01485&json=true","fetch_graph":"https://pith.science/api/pith-number/CHWLRHWPLSIHGYJBWS7CGBQJ3U/graph.json","fetch_events":"https://pith.science/api/pith-number/CHWLRHWPLSIHGYJBWS7CGBQJ3U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U/action/storage_attestation","attest_author":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U/action/author_attestation","sign_citation":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U/action/citation_signature","submit_replication":"https://pith.science/pith/CHWLRHWPLSIHGYJBWS7CGBQJ3U/action/replication_record"}},"created_at":"2026-07-05T08:23:19.367722+00:00","updated_at":"2026-07-05T08:23:19.367722+00:00"}