{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:HM4DGMNPWF4FPMAQBTERGSGM4W","short_pith_number":"pith:HM4DGMNP","schema_version":"1.0","canonical_sha256":"3b383331afb17857b0100cc91348cce59010fbe4de0a26fdfcda3aeb6f4b2536","source":{"kind":"arxiv","id":"1910.09792","version":3},"attestation_state":"computed","paper":{"title":"Robust Training with Ensemble Consensus","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jisoo Lee, Sae-Young Chung","submitted_at":"2019-10-22T06:58:10Z","abstract_excerpt":"Since deep neural networks are over-parameterized, they can memorize noisy examples. We address such a memorization issue in the presence of label noise. From the fact that deep neural networks cannot generalize to neighborhoods of memorized features, we hypothesize that noisy examples do not consistently incur small losses on the network under a certain perturbation. Based on this, we propose a novel training method called Learning with Ensemble Consensus (LEC) that prevents overfitting to noisy examples by removing them based on the consensus of an ensemble of perturbed networks. One of the "},"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":"1910.09792","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-10-22T06:58:10Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"c5efe53d988bdc950f1c46dd549c8754f0fd573273df93a7bfc74441d100f4cd","abstract_canon_sha256":"2b57af783efc168f78213b9e2c68f418d8f377181ce74d6a065f9d94729f93bd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:50:57.234845Z","signature_b64":"8NPiizL64pNlXAsczyjz6ORUP8BhLudbru8VIr4MhDk7WoyiKFxPNLhF61r/nP1bv7anXXB4056UgiRkjMslAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3b383331afb17857b0100cc91348cce59010fbe4de0a26fdfcda3aeb6f4b2536","last_reissued_at":"2026-07-05T01:50:57.234495Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:50:57.234495Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Robust Training with Ensemble Consensus","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jisoo Lee, Sae-Young Chung","submitted_at":"2019-10-22T06:58:10Z","abstract_excerpt":"Since deep neural networks are over-parameterized, they can memorize noisy examples. We address such a memorization issue in the presence of label noise. From the fact that deep neural networks cannot generalize to neighborhoods of memorized features, we hypothesize that noisy examples do not consistently incur small losses on the network under a certain perturbation. Based on this, we propose a novel training method called Learning with Ensemble Consensus (LEC) that prevents overfitting to noisy examples by removing them based on the consensus of an ensemble of perturbed networks. One of the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.09792","kind":"arxiv","version":3},"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/1910.09792/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":"1910.09792","created_at":"2026-07-05T01:50:57.234551+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.09792v3","created_at":"2026-07-05T01:50:57.234551+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.09792","created_at":"2026-07-05T01:50:57.234551+00:00"},{"alias_kind":"pith_short_12","alias_value":"HM4DGMNPWF4F","created_at":"2026-07-05T01:50:57.234551+00:00"},{"alias_kind":"pith_short_16","alias_value":"HM4DGMNPWF4FPMAQ","created_at":"2026-07-05T01:50:57.234551+00:00"},{"alias_kind":"pith_short_8","alias_value":"HM4DGMNP","created_at":"2026-07-05T01:50:57.234551+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.19581","citing_title":"Guiding Noisy Label Conditional Diffusion Models with Score-based Discriminator Correction","ref_index":31,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W","json":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W.json","graph_json":"https://pith.science/api/pith-number/HM4DGMNPWF4FPMAQBTERGSGM4W/graph.json","events_json":"https://pith.science/api/pith-number/HM4DGMNPWF4FPMAQBTERGSGM4W/events.json","paper":"https://pith.science/paper/HM4DGMNP"},"agent_actions":{"view_html":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W","download_json":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W.json","view_paper":"https://pith.science/paper/HM4DGMNP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.09792&json=true","fetch_graph":"https://pith.science/api/pith-number/HM4DGMNPWF4FPMAQBTERGSGM4W/graph.json","fetch_events":"https://pith.science/api/pith-number/HM4DGMNPWF4FPMAQBTERGSGM4W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W/action/storage_attestation","attest_author":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W/action/author_attestation","sign_citation":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W/action/citation_signature","submit_replication":"https://pith.science/pith/HM4DGMNPWF4FPMAQBTERGSGM4W/action/replication_record"}},"created_at":"2026-07-05T01:50:57.234551+00:00","updated_at":"2026-07-05T01:50:57.234551+00:00"}