{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:AKCGUNFDZDCY7KJ46PLATGRICX","short_pith_number":"pith:AKCGUNFD","schema_version":"1.0","canonical_sha256":"02846a34a3c8c58fa93cf3d6099a2815e77094b0beb2b8c9b0e1a03f51de9a75","source":{"kind":"arxiv","id":"2310.02541","version":1},"attestation_state":"computed","paper":{"title":"Benign Overfitting and Grokking in ReLU Networks for XOR Cluster Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Gal Vardi, Spencer Frei, Wei Hu, Yutong Wang, Zhiwei Xu","submitted_at":"2023-10-04T02:50:34Z","abstract_excerpt":"Neural networks trained by gradient descent (GD) have exhibited a number of surprising generalization behaviors. First, they can achieve a perfect fit to noisy training data and still generalize near-optimally, showing that overfitting can sometimes be benign. Second, they can undergo a period of classical, harmful overfitting -- achieving a perfect fit to training data with near-random performance on test data -- before transitioning (\"grokking\") to near-optimal generalization later in training. In this work, we show that both of these phenomena provably occur in two-layer ReLU networks train"},"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":"2310.02541","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-04T02:50:34Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"28f5777ef5eaa2f7fc5e011b3f876824350fe0066b9a707bbe0493ed2c382da3","abstract_canon_sha256":"b6be9dfd068aebf8a543eda6ad26183a67434d26d6ed108458ba7b35e0578bbd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:56:58.902859Z","signature_b64":"ipN1Ds21uG9Km6Ajj02UCYB8SNU+XObTb1S5l/jOtYG7ddvfF/x5ZkKVJm18z2FqXVI7uibtwjuIjGk75R8PBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"02846a34a3c8c58fa93cf3d6099a2815e77094b0beb2b8c9b0e1a03f51de9a75","last_reissued_at":"2026-07-05T06:56:58.902379Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:56:58.902379Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Benign Overfitting and Grokking in ReLU Networks for XOR Cluster Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Gal Vardi, Spencer Frei, Wei Hu, Yutong Wang, Zhiwei Xu","submitted_at":"2023-10-04T02:50:34Z","abstract_excerpt":"Neural networks trained by gradient descent (GD) have exhibited a number of surprising generalization behaviors. First, they can achieve a perfect fit to noisy training data and still generalize near-optimally, showing that overfitting can sometimes be benign. Second, they can undergo a period of classical, harmful overfitting -- achieving a perfect fit to training data with near-random performance on test data -- before transitioning (\"grokking\") to near-optimal generalization later in training. In this work, we show that both of these phenomena provably occur in two-layer ReLU networks train"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.02541","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/2310.02541/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":"2310.02541","created_at":"2026-07-05T06:56:58.902444+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.02541v1","created_at":"2026-07-05T06:56:58.902444+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.02541","created_at":"2026-07-05T06:56:58.902444+00:00"},{"alias_kind":"pith_short_12","alias_value":"AKCGUNFDZDCY","created_at":"2026-07-05T06:56:58.902444+00:00"},{"alias_kind":"pith_short_16","alias_value":"AKCGUNFDZDCY7KJ4","created_at":"2026-07-05T06:56:58.902444+00:00"},{"alias_kind":"pith_short_8","alias_value":"AKCGUNFD","created_at":"2026-07-05T06:56:58.902444+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.18022","citing_title":"Unveiling Memorization-Generalization Coexistence: A Case Study on Arithmetic Tasks with Label Noise","ref_index":42,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX","json":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX.json","graph_json":"https://pith.science/api/pith-number/AKCGUNFDZDCY7KJ46PLATGRICX/graph.json","events_json":"https://pith.science/api/pith-number/AKCGUNFDZDCY7KJ46PLATGRICX/events.json","paper":"https://pith.science/paper/AKCGUNFD"},"agent_actions":{"view_html":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX","download_json":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX.json","view_paper":"https://pith.science/paper/AKCGUNFD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.02541&json=true","fetch_graph":"https://pith.science/api/pith-number/AKCGUNFDZDCY7KJ46PLATGRICX/graph.json","fetch_events":"https://pith.science/api/pith-number/AKCGUNFDZDCY7KJ46PLATGRICX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX/action/storage_attestation","attest_author":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX/action/author_attestation","sign_citation":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX/action/citation_signature","submit_replication":"https://pith.science/pith/AKCGUNFDZDCY7KJ46PLATGRICX/action/replication_record"}},"created_at":"2026-07-05T06:56:58.902444+00:00","updated_at":"2026-07-05T06:56:58.902444+00:00"}