{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:CBNLOJESSDIFUQIAJ2DPCEXSLN","short_pith_number":"pith:CBNLOJES","schema_version":"1.0","canonical_sha256":"105ab7249290d05a41004e86f112f25b7d627f05fb010fe743b5424071dcb8bb","source":{"kind":"arxiv","id":"2312.03804","version":2},"attestation_state":"computed","paper":{"title":"How Low Can You Go? Surfacing Prototypical In-Distribution Samples for Unsupervised Anomaly Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alexander Ziller, Daniel Rueckert, Felix Meissen, Georgios Kaissis, Johannes Getzner, Martin J. Menten, \\\"Ozg\\\"un Turgut","submitted_at":"2023-12-06T15:30:47Z","abstract_excerpt":"Unsupervised anomaly detection (UAD) alleviates large labeling efforts by training exclusively on unlabeled in-distribution data and detecting outliers as anomalies. Generally, the assumption prevails that large training datasets allow the training of higher-performing UAD models. However, in this work, we show that UAD with extremely few training samples can already match -- and in some cases even surpass -- the performance of training with the whole training dataset. Building upon this finding, we propose an unsupervised method to reliably identify prototypical samples to further boost UAD p"},"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":"2312.03804","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-12-06T15:30:47Z","cross_cats_sorted":[],"title_canon_sha256":"38b4b461053c1846ef4e0d78d11f486ebe098aebd7748ce110d2546f6a9cef46","abstract_canon_sha256":"22a501b2e13798a82ab5c03394377d28f2e0277ab064a408c5718c39255f0cd9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:26:57.294868Z","signature_b64":"v64JzwEaFOJmJL25gwsGsBEPA55uh/W/nQefu+YlIVyQ5T3IiDbxo1PWqj2Chr4SOSFV7TGXdEcKLB81S4poAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"105ab7249290d05a41004e86f112f25b7d627f05fb010fe743b5424071dcb8bb","last_reissued_at":"2026-07-05T09:26:57.294308Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:26:57.294308Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Low Can You Go? Surfacing Prototypical In-Distribution Samples for Unsupervised Anomaly Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alexander Ziller, Daniel Rueckert, Felix Meissen, Georgios Kaissis, Johannes Getzner, Martin J. Menten, \\\"Ozg\\\"un Turgut","submitted_at":"2023-12-06T15:30:47Z","abstract_excerpt":"Unsupervised anomaly detection (UAD) alleviates large labeling efforts by training exclusively on unlabeled in-distribution data and detecting outliers as anomalies. Generally, the assumption prevails that large training datasets allow the training of higher-performing UAD models. However, in this work, we show that UAD with extremely few training samples can already match -- and in some cases even surpass -- the performance of training with the whole training dataset. Building upon this finding, we propose an unsupervised method to reliably identify prototypical samples to further boost UAD p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.03804","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/2312.03804/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":"2312.03804","created_at":"2026-07-05T09:26:57.294384+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.03804v2","created_at":"2026-07-05T09:26:57.294384+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.03804","created_at":"2026-07-05T09:26:57.294384+00:00"},{"alias_kind":"pith_short_12","alias_value":"CBNLOJESSDIF","created_at":"2026-07-05T09:26:57.294384+00:00"},{"alias_kind":"pith_short_16","alias_value":"CBNLOJESSDIFUQIA","created_at":"2026-07-05T09:26:57.294384+00:00"},{"alias_kind":"pith_short_8","alias_value":"CBNLOJES","created_at":"2026-07-05T09:26:57.294384+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN","json":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN.json","graph_json":"https://pith.science/api/pith-number/CBNLOJESSDIFUQIAJ2DPCEXSLN/graph.json","events_json":"https://pith.science/api/pith-number/CBNLOJESSDIFUQIAJ2DPCEXSLN/events.json","paper":"https://pith.science/paper/CBNLOJES"},"agent_actions":{"view_html":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN","download_json":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN.json","view_paper":"https://pith.science/paper/CBNLOJES","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.03804&json=true","fetch_graph":"https://pith.science/api/pith-number/CBNLOJESSDIFUQIAJ2DPCEXSLN/graph.json","fetch_events":"https://pith.science/api/pith-number/CBNLOJESSDIFUQIAJ2DPCEXSLN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN/action/storage_attestation","attest_author":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN/action/author_attestation","sign_citation":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN/action/citation_signature","submit_replication":"https://pith.science/pith/CBNLOJESSDIFUQIAJ2DPCEXSLN/action/replication_record"}},"created_at":"2026-07-05T09:26:57.294384+00:00","updated_at":"2026-07-05T09:26:57.294384+00:00"}