{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:SC32Z4NB7GRQDIB3DB3MRIXLGD","short_pith_number":"pith:SC32Z4NB","schema_version":"1.0","canonical_sha256":"90b7acf1a1f9a301a03b1876c8a2eb30f5e3e9587183ae6fed2ddb1d3d70457d","source":{"kind":"arxiv","id":"2410.13831","version":2},"attestation_state":"computed","paper":{"title":"The Disparate Benefits of Deep Ensembles","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Adrian Arnaiz-Rodriguez, Kajetan Schweighofer, Nuria Oliver, Sepp Hochreiter","submitted_at":"2024-10-17T17:53:01Z","abstract_excerpt":"Ensembles of Deep Neural Networks, Deep Ensembles, are widely used as a simple way to boost predictive performance. However, their impact on algorithmic fairness is not well understood yet. Algorithmic fairness examines how a model's performance varies across socially relevant groups defined by protected attributes such as age, gender, or race. In this work, we explore the interplay between the performance gains from Deep Ensembles and fairness. Our analysis reveals that they unevenly favor different groups, a phenomenon that we term the disparate benefits effect. We empirically investigate th"},"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":"2410.13831","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-17T17:53:01Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0134e5632e22218ee91c0164ee2695128716d436f0e28a58a9d57a45adef269b","abstract_canon_sha256":"84cb913ae51f3dde43d5217280c20e41c0f6f865e0bc7a8ce097f0a17caa50e1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:04.143077Z","signature_b64":"DosrV+Rz0yy2CMoE9XUfOp4/Dj4W3fwlf31+Wbax5o/6Rnp94OjfqDS1picwGM3eDMmwiQr4LHobL9/OM/qgCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"90b7acf1a1f9a301a03b1876c8a2eb30f5e3e9587183ae6fed2ddb1d3d70457d","last_reissued_at":"2026-07-05T11:16:04.142583Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:04.142583Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Disparate Benefits of Deep Ensembles","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Adrian Arnaiz-Rodriguez, Kajetan Schweighofer, Nuria Oliver, Sepp Hochreiter","submitted_at":"2024-10-17T17:53:01Z","abstract_excerpt":"Ensembles of Deep Neural Networks, Deep Ensembles, are widely used as a simple way to boost predictive performance. However, their impact on algorithmic fairness is not well understood yet. Algorithmic fairness examines how a model's performance varies across socially relevant groups defined by protected attributes such as age, gender, or race. In this work, we explore the interplay between the performance gains from Deep Ensembles and fairness. Our analysis reveals that they unevenly favor different groups, a phenomenon that we term the disparate benefits effect. We empirically investigate th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13831","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/2410.13831/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":"2410.13831","created_at":"2026-07-05T11:16:04.142634+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.13831v2","created_at":"2026-07-05T11:16:04.142634+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13831","created_at":"2026-07-05T11:16:04.142634+00:00"},{"alias_kind":"pith_short_12","alias_value":"SC32Z4NB7GRQ","created_at":"2026-07-05T11:16:04.142634+00:00"},{"alias_kind":"pith_short_16","alias_value":"SC32Z4NB7GRQDIB3","created_at":"2026-07-05T11:16:04.142634+00:00"},{"alias_kind":"pith_short_8","alias_value":"SC32Z4NB","created_at":"2026-07-05T11:16:04.142634+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.14551","citing_title":"Fairness of Deep Ensembles: On the interplay between per-group task difficulty and under-representation","ref_index":26,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD","json":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD.json","graph_json":"https://pith.science/api/pith-number/SC32Z4NB7GRQDIB3DB3MRIXLGD/graph.json","events_json":"https://pith.science/api/pith-number/SC32Z4NB7GRQDIB3DB3MRIXLGD/events.json","paper":"https://pith.science/paper/SC32Z4NB"},"agent_actions":{"view_html":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD","download_json":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD.json","view_paper":"https://pith.science/paper/SC32Z4NB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.13831&json=true","fetch_graph":"https://pith.science/api/pith-number/SC32Z4NB7GRQDIB3DB3MRIXLGD/graph.json","fetch_events":"https://pith.science/api/pith-number/SC32Z4NB7GRQDIB3DB3MRIXLGD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD/action/storage_attestation","attest_author":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD/action/author_attestation","sign_citation":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD/action/citation_signature","submit_replication":"https://pith.science/pith/SC32Z4NB7GRQDIB3DB3MRIXLGD/action/replication_record"}},"created_at":"2026-07-05T11:16:04.142634+00:00","updated_at":"2026-07-05T11:16:04.142634+00:00"}