{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:VTYTMET2WS25J6E5KEEZRNEIZY","short_pith_number":"pith:VTYTMET2","schema_version":"1.0","canonical_sha256":"acf136127ab4b5d4f89d510998b488ce22716d46eb0e4dbfb86b9bc9df1127c0","source":{"kind":"arxiv","id":"2010.13132","version":3},"attestation_state":"computed","paper":{"title":"Multiscale Score Matching for Out-of-Distribution Detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ahsan Mahmood, Junier Oliva, Martin Styner","submitted_at":"2020-10-25T15:10:31Z","abstract_excerpt":"We present a new methodology for detecting out-of-distribution (OOD) images by utilizing norms of the score estimates at multiple noise scales. A score is defined to be the gradient of the log density with respect to the input data. Our methodology is completely unsupervised and follows a straight forward training scheme. First, we train a deep network to estimate scores for levels of noise. Once trained, we calculate the noisy score estimates for N in-distribution samples and take the L2-norms across the input dimensions (resulting in an NxL matrix). Then we train an auxiliary model (such as "},"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":"2010.13132","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-10-25T15:10:31Z","cross_cats_sorted":[],"title_canon_sha256":"ec7efdb8c2ecad23030f225ab0e5bc2d5848ae6a89a2efe1fb60b1661b4d3e5b","abstract_canon_sha256":"531bd43c3d12e0d67eb9be2319714033f8c73b3cb2c1c41f15ac7155972a9e1f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:25:27.324308Z","signature_b64":"KMpVmc7HniSEuZA4sv41VInA9C6u/+LQ5xPx8kMGm76N3/dQrnP5VArjwPDdpIpwnNOiDp1TeuhoJq9gr4fnCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"acf136127ab4b5d4f89d510998b488ce22716d46eb0e4dbfb86b9bc9df1127c0","last_reissued_at":"2026-07-05T02:25:27.323840Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:25:27.323840Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multiscale Score Matching for Out-of-Distribution Detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ahsan Mahmood, Junier Oliva, Martin Styner","submitted_at":"2020-10-25T15:10:31Z","abstract_excerpt":"We present a new methodology for detecting out-of-distribution (OOD) images by utilizing norms of the score estimates at multiple noise scales. A score is defined to be the gradient of the log density with respect to the input data. Our methodology is completely unsupervised and follows a straight forward training scheme. First, we train a deep network to estimate scores for levels of noise. Once trained, we calculate the noisy score estimates for N in-distribution samples and take the L2-norms across the input dimensions (resulting in an NxL matrix). Then we train an auxiliary model (such as "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.13132","kind":"arxiv","version":3},"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/2010.13132/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":"2010.13132","created_at":"2026-07-05T02:25:27.323894+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.13132v3","created_at":"2026-07-05T02:25:27.323894+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.13132","created_at":"2026-07-05T02:25:27.323894+00:00"},{"alias_kind":"pith_short_12","alias_value":"VTYTMET2WS25","created_at":"2026-07-05T02:25:27.323894+00:00"},{"alias_kind":"pith_short_16","alias_value":"VTYTMET2WS25J6E5","created_at":"2026-07-05T02:25:27.323894+00:00"},{"alias_kind":"pith_short_8","alias_value":"VTYTMET2","created_at":"2026-07-05T02:25:27.323894+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.25080","citing_title":"Towards a Certificate of Trust: Task-Aware OOD Detection for Scientific AI","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05638","citing_title":"Scaling Pretrained Representations Enables Label-Free Out-of-Distribution Detection Without Fine-Tuning","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14914","citing_title":"Beyond Prompts: Unconditional 3D Inversion for Out-of-Distribution Shapes","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY","json":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY.json","graph_json":"https://pith.science/api/pith-number/VTYTMET2WS25J6E5KEEZRNEIZY/graph.json","events_json":"https://pith.science/api/pith-number/VTYTMET2WS25J6E5KEEZRNEIZY/events.json","paper":"https://pith.science/paper/VTYTMET2"},"agent_actions":{"view_html":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY","download_json":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY.json","view_paper":"https://pith.science/paper/VTYTMET2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.13132&json=true","fetch_graph":"https://pith.science/api/pith-number/VTYTMET2WS25J6E5KEEZRNEIZY/graph.json","fetch_events":"https://pith.science/api/pith-number/VTYTMET2WS25J6E5KEEZRNEIZY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY/action/storage_attestation","attest_author":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY/action/author_attestation","sign_citation":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY/action/citation_signature","submit_replication":"https://pith.science/pith/VTYTMET2WS25J6E5KEEZRNEIZY/action/replication_record"}},"created_at":"2026-07-05T02:25:27.323894+00:00","updated_at":"2026-07-05T02:25:27.323894+00:00"}