{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:N6NEEHOAEUUJFZYY5BDPCSR5K3","short_pith_number":"pith:N6NEEHOA","schema_version":"1.0","canonical_sha256":"6f9a421dc0252892e718e846f14a3d56fac63ed95f98d55c3c22d1970efb93db","source":{"kind":"arxiv","id":"2402.08946","version":1},"attestation_state":"computed","paper":{"title":"Measuring Sharpness in Grokking","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Charles O'Neill, Jack Miller, Noam Levi, Patrick Gleeson, Thang Bui","submitted_at":"2024-02-14T05:22:53Z","abstract_excerpt":"Neural networks sometimes exhibit grokking, a phenomenon where perfect or near-perfect performance is achieved on a validation set well after the same performance has been obtained on the corresponding training set. In this workshop paper, we introduce a robust technique for measuring grokking, based on fitting an appropriate functional form. We then use this to investigate the sharpness of transitions in training and validation accuracy under two settings. The first setting is the theoretical framework developed by Levi et al. (2023) where closed form expressions are readily accessible. The s"},"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":"2402.08946","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-02-14T05:22:53Z","cross_cats_sorted":[],"title_canon_sha256":"0cc7e98a4007a71609c01aef147fca2310f870fe0528eb9679cee47ff436a6b6","abstract_canon_sha256":"9ca69822ba4e7c8284f619e02ea1a824e2a1cce776691f1ce83d3a0f7c872f08"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:45:05.903574Z","signature_b64":"YkqF9Crum/SoM//jTQZRCsUoznFwEy2HnlrlyxTGim5WxtZuANJoyjSPPgCCWcKGtiJQ1a/JFiowZxwIsm0kDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6f9a421dc0252892e718e846f14a3d56fac63ed95f98d55c3c22d1970efb93db","last_reissued_at":"2026-07-05T07:45:05.903046Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:45:05.903046Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Measuring Sharpness in Grokking","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Charles O'Neill, Jack Miller, Noam Levi, Patrick Gleeson, Thang Bui","submitted_at":"2024-02-14T05:22:53Z","abstract_excerpt":"Neural networks sometimes exhibit grokking, a phenomenon where perfect or near-perfect performance is achieved on a validation set well after the same performance has been obtained on the corresponding training set. In this workshop paper, we introduce a robust technique for measuring grokking, based on fitting an appropriate functional form. We then use this to investigate the sharpness of transitions in training and validation accuracy under two settings. The first setting is the theoretical framework developed by Levi et al. (2023) where closed form expressions are readily accessible. The s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.08946","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/2402.08946/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":"2402.08946","created_at":"2026-07-05T07:45:05.903111+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.08946v1","created_at":"2026-07-05T07:45:05.903111+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.08946","created_at":"2026-07-05T07:45:05.903111+00:00"},{"alias_kind":"pith_short_12","alias_value":"N6NEEHOAEUUJ","created_at":"2026-07-05T07:45:05.903111+00:00"},{"alias_kind":"pith_short_16","alias_value":"N6NEEHOAEUUJFZYY","created_at":"2026-07-05T07:45:05.903111+00:00"},{"alias_kind":"pith_short_8","alias_value":"N6NEEHOA","created_at":"2026-07-05T07:45:05.903111+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06639","citing_title":"At-Grok Is Not Converged:A Measurement-Validity Audit for Grokking Representation Metrics","ref_index":10,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3","json":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3.json","graph_json":"https://pith.science/api/pith-number/N6NEEHOAEUUJFZYY5BDPCSR5K3/graph.json","events_json":"https://pith.science/api/pith-number/N6NEEHOAEUUJFZYY5BDPCSR5K3/events.json","paper":"https://pith.science/paper/N6NEEHOA"},"agent_actions":{"view_html":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3","download_json":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3.json","view_paper":"https://pith.science/paper/N6NEEHOA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.08946&json=true","fetch_graph":"https://pith.science/api/pith-number/N6NEEHOAEUUJFZYY5BDPCSR5K3/graph.json","fetch_events":"https://pith.science/api/pith-number/N6NEEHOAEUUJFZYY5BDPCSR5K3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3/action/storage_attestation","attest_author":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3/action/author_attestation","sign_citation":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3/action/citation_signature","submit_replication":"https://pith.science/pith/N6NEEHOAEUUJFZYY5BDPCSR5K3/action/replication_record"}},"created_at":"2026-07-05T07:45:05.903111+00:00","updated_at":"2026-07-05T07:45:05.903111+00:00"}