{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:LJRWQJ6ARWBRT7F5K2AGNDC452","short_pith_number":"pith:LJRWQJ6A","schema_version":"1.0","canonical_sha256":"5a636827c08d8319fcbd5680668c5ceebb1f383430786bc8e3f0afdaa6dec934","source":{"kind":"arxiv","id":"2407.10633","version":1},"attestation_state":"computed","paper":{"title":"Evaluating Model Bias Requires Characterizing its Mistakes","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"David Warde-Farley, Isabela Albuquerque, Jessica Schrouff, Olivia Wiles, Sven Gowal, Taylan Cemgil","submitted_at":"2024-07-15T11:46:21Z","abstract_excerpt":"The ability to properly benchmark model performance in the face of spurious correlations is important to both build better predictors and increase confidence that models are operating as intended. We demonstrate that characterizing (as opposed to simply quantifying) model mistakes across subgroups is pivotal to properly reflect model biases, which are ignored by standard metrics such as worst-group accuracy or accuracy gap. Inspired by the hypothesis testing framework, we introduce SkewSize, a principled and flexible metric that captures bias from mistakes in a model's predictions. It can be u"},"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":"2407.10633","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-07-15T11:46:21Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"d4859b2540e36125a8b9f720da2fdf3bf6d42c830ec0f51e0b2a5193159b9492","abstract_canon_sha256":"46424c92ce59fb9dcfb08052e8efe9e006500106ae6e0f36c5d7382ae4998f7e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:44:02.298099Z","signature_b64":"uQi84XGcjVODUm79tUvZi+u+kKXww8ra+KG+3FYUcgdxw0bjFc0RRbU4X/eNaJvHFPo5Wf07/uzauwMFobOJAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a636827c08d8319fcbd5680668c5ceebb1f383430786bc8e3f0afdaa6dec934","last_reissued_at":"2026-07-05T08:44:02.297698Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:44:02.297698Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluating Model Bias Requires Characterizing its Mistakes","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"David Warde-Farley, Isabela Albuquerque, Jessica Schrouff, Olivia Wiles, Sven Gowal, Taylan Cemgil","submitted_at":"2024-07-15T11:46:21Z","abstract_excerpt":"The ability to properly benchmark model performance in the face of spurious correlations is important to both build better predictors and increase confidence that models are operating as intended. We demonstrate that characterizing (as opposed to simply quantifying) model mistakes across subgroups is pivotal to properly reflect model biases, which are ignored by standard metrics such as worst-group accuracy or accuracy gap. Inspired by the hypothesis testing framework, we introduce SkewSize, a principled and flexible metric that captures bias from mistakes in a model's predictions. It can be u"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.10633","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/2407.10633/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":"2407.10633","created_at":"2026-07-05T08:44:02.297757+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.10633v1","created_at":"2026-07-05T08:44:02.297757+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.10633","created_at":"2026-07-05T08:44:02.297757+00:00"},{"alias_kind":"pith_short_12","alias_value":"LJRWQJ6ARWBR","created_at":"2026-07-05T08:44:02.297757+00:00"},{"alias_kind":"pith_short_16","alias_value":"LJRWQJ6ARWBRT7F5","created_at":"2026-07-05T08:44:02.297757+00:00"},{"alias_kind":"pith_short_8","alias_value":"LJRWQJ6A","created_at":"2026-07-05T08:44:02.297757+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/LJRWQJ6ARWBRT7F5K2AGNDC452","json":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452.json","graph_json":"https://pith.science/api/pith-number/LJRWQJ6ARWBRT7F5K2AGNDC452/graph.json","events_json":"https://pith.science/api/pith-number/LJRWQJ6ARWBRT7F5K2AGNDC452/events.json","paper":"https://pith.science/paper/LJRWQJ6A"},"agent_actions":{"view_html":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452","download_json":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452.json","view_paper":"https://pith.science/paper/LJRWQJ6A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.10633&json=true","fetch_graph":"https://pith.science/api/pith-number/LJRWQJ6ARWBRT7F5K2AGNDC452/graph.json","fetch_events":"https://pith.science/api/pith-number/LJRWQJ6ARWBRT7F5K2AGNDC452/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452/action/storage_attestation","attest_author":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452/action/author_attestation","sign_citation":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452/action/citation_signature","submit_replication":"https://pith.science/pith/LJRWQJ6ARWBRT7F5K2AGNDC452/action/replication_record"}},"created_at":"2026-07-05T08:44:02.297757+00:00","updated_at":"2026-07-05T08:44:02.297757+00:00"}