{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:PLZFHA3P7IMD6LHZAXPET6TLQE","short_pith_number":"pith:PLZFHA3P","schema_version":"1.0","canonical_sha256":"7af253836ffa183f2cf905de49fa6b8133f65f0d59ea4568a665ee9e25af584b","source":{"kind":"arxiv","id":"2110.10832","version":4},"attestation_state":"computed","paper":{"title":"Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Caiming Xiong, Devansh Arpit, Huan Wang, Yingbo Zhou","submitted_at":"2021-10-21T00:08:17Z","abstract_excerpt":"In Domain Generalization (DG) settings, models trained independently on a given set of training domains have notoriously chaotic performance on distribution shifted test domains, and stochasticity in optimization (e.g. seed) plays a big role. This makes deep learning models unreliable in real world settings. We first show that this chaotic behavior exists even along the training optimization trajectory of a single model, and propose a simple model averaging protocol that both significantly boosts domain generalization and diminishes the impact of stochasticity by improving the rank correlation"},"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":"2110.10832","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-10-21T00:08:17Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"cc20b776cec7d0d85d8fb67df1945401d2f73b6b12facdde861fb5a8bef52777","abstract_canon_sha256":"35ebdbea446848c55c386126ea0afe559409718fe87327baa81e2c0217807790"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:06:08.695381Z","signature_b64":"3po2lYqhdgbSFp50EaTTzjgDJfD3nmtIj+n5zpoAf5VXCuyy9wLIK3xkqvarG+77xLJ0pi2xXgoFt1BiKTmUCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7af253836ffa183f2cf905de49fa6b8133f65f0d59ea4568a665ee9e25af584b","last_reissued_at":"2026-07-05T05:06:08.694846Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:06:08.694846Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Caiming Xiong, Devansh Arpit, Huan Wang, Yingbo Zhou","submitted_at":"2021-10-21T00:08:17Z","abstract_excerpt":"In Domain Generalization (DG) settings, models trained independently on a given set of training domains have notoriously chaotic performance on distribution shifted test domains, and stochasticity in optimization (e.g. seed) plays a big role. This makes deep learning models unreliable in real world settings. We first show that this chaotic behavior exists even along the training optimization trajectory of a single model, and propose a simple model averaging protocol that both significantly boosts domain generalization and diminishes the impact of stochasticity by improving the rank correlation"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.10832","kind":"arxiv","version":4},"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/2110.10832/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":"2110.10832","created_at":"2026-07-05T05:06:08.694911+00:00"},{"alias_kind":"arxiv_version","alias_value":"2110.10832v4","created_at":"2026-07-05T05:06:08.694911+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.10832","created_at":"2026-07-05T05:06:08.694911+00:00"},{"alias_kind":"pith_short_12","alias_value":"PLZFHA3P7IMD","created_at":"2026-07-05T05:06:08.694911+00:00"},{"alias_kind":"pith_short_16","alias_value":"PLZFHA3P7IMD6LHZ","created_at":"2026-07-05T05:06:08.694911+00:00"},{"alias_kind":"pith_short_8","alias_value":"PLZFHA3P","created_at":"2026-07-05T05:06:08.694911+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2209.14742","citing_title":"Learning Gradient-based Mixup with Extrapolation toward Flatter Minima for Domain Generalization","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2502.05564","citing_title":"TabICL: A Tabular Foundation Model for In-Context Learning on Large Data","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE","json":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE.json","graph_json":"https://pith.science/api/pith-number/PLZFHA3P7IMD6LHZAXPET6TLQE/graph.json","events_json":"https://pith.science/api/pith-number/PLZFHA3P7IMD6LHZAXPET6TLQE/events.json","paper":"https://pith.science/paper/PLZFHA3P"},"agent_actions":{"view_html":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE","download_json":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE.json","view_paper":"https://pith.science/paper/PLZFHA3P","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2110.10832&json=true","fetch_graph":"https://pith.science/api/pith-number/PLZFHA3P7IMD6LHZAXPET6TLQE/graph.json","fetch_events":"https://pith.science/api/pith-number/PLZFHA3P7IMD6LHZAXPET6TLQE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE/action/storage_attestation","attest_author":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE/action/author_attestation","sign_citation":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE/action/citation_signature","submit_replication":"https://pith.science/pith/PLZFHA3P7IMD6LHZAXPET6TLQE/action/replication_record"}},"created_at":"2026-07-05T05:06:08.694911+00:00","updated_at":"2026-07-05T05:06:08.694911+00:00"}