{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7YCRD3ZO3QRCHONI6JSYWXPB4N","short_pith_number":"pith:7YCRD3ZO","schema_version":"1.0","canonical_sha256":"fe0511ef2edc2223b9a8f2658b5de1e34fda6965baf89208245b4505e69faf14","source":{"kind":"arxiv","id":"2605.18329","version":1},"attestation_state":"computed","paper":{"title":"Lost in the Folds: When Cross-Validation Is Not a Deep Ensemble for Uncertainty Estimation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Bujotzek Markus (DKFZ), DKFZ), Isensee Fabian (DKFZ), Kahl Kim-Celine (DKFZ), Kirchhoff Yannick (DKFZ), Kirscher Tristan (ICube, Kovacs Balint (DKFZ), Maier-Hein Klaus (DKFZ), Rokuss Maximilian (DKFZ)","submitted_at":"2026-05-18T12:46:57Z","abstract_excerpt":"Ensemble disagreement is widely used as a proxy for epistemic uncertainty in medical image segmentation. In practice, many studies form ensembles via K-fold cross-validation (CV), yet refer to them as ``deep ensembles'' (DE). Because CV members are trained on different data subsets, their disagreement mixes seed-driven variability with data-exposure effects, which can change how uncertainty should be interpreted. We audit recent segmentation uncertainty studies and find that terminology--implementation mismatches are common. We then compare a standard 5-fold CV ensemble to a 5-member DE (fixed"},"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":"2605.18329","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-18T12:46:57Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"071c411e48fe30b202a5f8249f388bcc80fca8b0732923f8d3d876db97e929da","abstract_canon_sha256":"cdfc47fff36840d5e0069e9f89e7583c992c883f7adfd60059815e6328a3f5ce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:05:55.503659Z","signature_b64":"YHZYY4JEVzGyLN2x560ikq6vr/9Yily2yWJiI3WvE1qmkk19rMKzWogNquv4UqBeOddegTW65m3aMkSOC2tEDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fe0511ef2edc2223b9a8f2658b5de1e34fda6965baf89208245b4505e69faf14","last_reissued_at":"2026-05-20T00:05:55.502938Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:05:55.502938Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Lost in the Folds: When Cross-Validation Is Not a Deep Ensemble for Uncertainty Estimation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Bujotzek Markus (DKFZ), DKFZ), Isensee Fabian (DKFZ), Kahl Kim-Celine (DKFZ), Kirchhoff Yannick (DKFZ), Kirscher Tristan (ICube, Kovacs Balint (DKFZ), Maier-Hein Klaus (DKFZ), Rokuss Maximilian (DKFZ)","submitted_at":"2026-05-18T12:46:57Z","abstract_excerpt":"Ensemble disagreement is widely used as a proxy for epistemic uncertainty in medical image segmentation. In practice, many studies form ensembles via K-fold cross-validation (CV), yet refer to them as ``deep ensembles'' (DE). Because CV members are trained on different data subsets, their disagreement mixes seed-driven variability with data-exposure effects, which can change how uncertainty should be interpreted. We audit recent segmentation uncertainty studies and find that terminology--implementation mismatches are common. We then compare a standard 5-fold CV ensemble to a 5-member DE (fixed"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.18329","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/2605.18329/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-19T23:33:35.178341Z","status":"skipped","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T23:21:58.848615Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"407d0d265cc639bfd4eee6bc9451ed15300fbd147326eb7863c0f9bf40a96a2a"},"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":"2605.18329","created_at":"2026-05-20T00:05:55.503047+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.18329v1","created_at":"2026-05-20T00:05:55.503047+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.18329","created_at":"2026-05-20T00:05:55.503047+00:00"},{"alias_kind":"pith_short_12","alias_value":"7YCRD3ZO3QRC","created_at":"2026-05-20T00:05:55.503047+00:00"},{"alias_kind":"pith_short_16","alias_value":"7YCRD3ZO3QRCHONI","created_at":"2026-05-20T00:05:55.503047+00:00"},{"alias_kind":"pith_short_8","alias_value":"7YCRD3ZO","created_at":"2026-05-20T00:05:55.503047+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/7YCRD3ZO3QRCHONI6JSYWXPB4N","json":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N.json","graph_json":"https://pith.science/api/pith-number/7YCRD3ZO3QRCHONI6JSYWXPB4N/graph.json","events_json":"https://pith.science/api/pith-number/7YCRD3ZO3QRCHONI6JSYWXPB4N/events.json","paper":"https://pith.science/paper/7YCRD3ZO"},"agent_actions":{"view_html":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N","download_json":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N.json","view_paper":"https://pith.science/paper/7YCRD3ZO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.18329&json=true","fetch_graph":"https://pith.science/api/pith-number/7YCRD3ZO3QRCHONI6JSYWXPB4N/graph.json","fetch_events":"https://pith.science/api/pith-number/7YCRD3ZO3QRCHONI6JSYWXPB4N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N/action/storage_attestation","attest_author":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N/action/author_attestation","sign_citation":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N/action/citation_signature","submit_replication":"https://pith.science/pith/7YCRD3ZO3QRCHONI6JSYWXPB4N/action/replication_record"}},"created_at":"2026-05-20T00:05:55.503047+00:00","updated_at":"2026-05-20T00:05:55.503047+00:00"}