{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NJQYVUXQRL2JEYJKS34OMJW2T5","short_pith_number":"pith:NJQYVUXQ","schema_version":"1.0","canonical_sha256":"6a618ad2f08af492612a96f8e626da9f58cdec88d56b4db2cf2a603ee19fea76","source":{"kind":"arxiv","id":"2412.07684","version":1},"attestation_state":"computed","paper":{"title":"The Pitfalls of Memorization: When Memorization Hurts Generalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"David Lopez-Paz, Elvis Dohmatob, Mohammad Pezeshki, Pascal Vincent, Reza Bayat","submitted_at":"2024-12-10T17:18:33Z","abstract_excerpt":"Neural networks often learn simple explanations that fit the majority of the data while memorizing exceptions that deviate from these explanations.This behavior leads to poor generalization when the learned explanations rely on spurious correlations. In this work, we formalize the interplay between memorization and generalization, showing that spurious correlations would particularly lead to poor generalization when are combined with memorization. Memorization can reduce training loss to zero, leaving no incentive to learn robust, generalizable patterns. To address this, we propose memorizatio"},"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":"2412.07684","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-10T17:18:33Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"145ad3a364104d3a799189fc208f3b49c355e201c953f896bcbf6506f8b726d3","abstract_canon_sha256":"3082e3d1f5779b1ff8488a934559daee72ea36ddab8576c544fe7abed5186fb7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:47:13.664359Z","signature_b64":"rTyWtCBsUJs0B6eHXAZVgRn1Eeqsxw4dieLzE4pzD5XRQr0lhxzNoOh2pLvc+doeNcdbbE7I0hhuRUsm9CibDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6a618ad2f08af492612a96f8e626da9f58cdec88d56b4db2cf2a603ee19fea76","last_reissued_at":"2026-07-05T09:47:13.663841Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:47:13.663841Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Pitfalls of Memorization: When Memorization Hurts Generalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"David Lopez-Paz, Elvis Dohmatob, Mohammad Pezeshki, Pascal Vincent, Reza Bayat","submitted_at":"2024-12-10T17:18:33Z","abstract_excerpt":"Neural networks often learn simple explanations that fit the majority of the data while memorizing exceptions that deviate from these explanations.This behavior leads to poor generalization when the learned explanations rely on spurious correlations. In this work, we formalize the interplay between memorization and generalization, showing that spurious correlations would particularly lead to poor generalization when are combined with memorization. Memorization can reduce training loss to zero, leaving no incentive to learn robust, generalizable patterns. To address this, we propose memorizatio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.07684","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/2412.07684/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":"2412.07684","created_at":"2026-07-05T09:47:13.663904+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.07684v1","created_at":"2026-07-05T09:47:13.663904+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.07684","created_at":"2026-07-05T09:47:13.663904+00:00"},{"alias_kind":"pith_short_12","alias_value":"NJQYVUXQRL2J","created_at":"2026-07-05T09:47:13.663904+00:00"},{"alias_kind":"pith_short_16","alias_value":"NJQYVUXQRL2JEYJK","created_at":"2026-07-05T09:47:13.663904+00:00"},{"alias_kind":"pith_short_8","alias_value":"NJQYVUXQ","created_at":"2026-07-05T09:47:13.663904+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.02830","citing_title":"Mitigating Spurious Correlations with Memorization-Guided Dataset De-Biasing","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2501.00663","citing_title":"Titans: Learning to Memorize at Test Time","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11170","citing_title":"Unlearning with Asymmetric Sources: Improved Unlearning-Utility Trade-off with Public Data","ref_index":118,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5","json":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5.json","graph_json":"https://pith.science/api/pith-number/NJQYVUXQRL2JEYJKS34OMJW2T5/graph.json","events_json":"https://pith.science/api/pith-number/NJQYVUXQRL2JEYJKS34OMJW2T5/events.json","paper":"https://pith.science/paper/NJQYVUXQ"},"agent_actions":{"view_html":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5","download_json":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5.json","view_paper":"https://pith.science/paper/NJQYVUXQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.07684&json=true","fetch_graph":"https://pith.science/api/pith-number/NJQYVUXQRL2JEYJKS34OMJW2T5/graph.json","fetch_events":"https://pith.science/api/pith-number/NJQYVUXQRL2JEYJKS34OMJW2T5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5/action/storage_attestation","attest_author":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5/action/author_attestation","sign_citation":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5/action/citation_signature","submit_replication":"https://pith.science/pith/NJQYVUXQRL2JEYJKS34OMJW2T5/action/replication_record"}},"created_at":"2026-07-05T09:47:13.663904+00:00","updated_at":"2026-07-05T09:47:13.663904+00:00"}