{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:JQX2LNR4MW6DTPSZI46444CLAS","short_pith_number":"pith:JQX2LNR4","schema_version":"1.0","canonical_sha256":"4c2fa5b63c65bc39be59473dce704b049aa64d6851eb3fdc9278a40e22efbee7","source":{"kind":"arxiv","id":"2607.11666","version":1},"attestation_state":"computed","paper":{"title":"How to Tame Grokking: Representation Geometry as a Control Signal","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Maksim A Kazanskii","submitted_at":"2026-07-13T15:09:42Z","abstract_excerpt":"Grokking is a phenomenon in which neural networks initially memorize training data and only later exhibit strong generalization after prolonged optimization. Despite extensive recent study, the factors influencing the emergence and timing of grokking remain incompletely understood. We investigate the relationship between representation geometry and delayed generalization. We find that dimensionality collapse consistently precedes the onset of grokking in all evaluated settings. Motivated by these observations, we introduce Geometric Dimensionality Regularization (GeomDR), a simple spectral reg"},"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":"2607.11666","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-13T15:09:42Z","cross_cats_sorted":[],"title_canon_sha256":"a0e89f1ce0e8e25bc69954c3cb3ede9c278713158df01ed1e386f22ebe458c49","abstract_canon_sha256":"bb33193f20d81c3ea326cdb013f6b6537620141cf8293563ad83bd89aa9a3586"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-14T02:22:17.585426Z","signature_b64":"9IXyJVxGVZH5c2p6qT7QR+hV4qXfpLVopEHJdgk9mPdrpwnVZVSTtmQCk5MIFJzxiEr6CLBT69PPJ8ptqEwNDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4c2fa5b63c65bc39be59473dce704b049aa64d6851eb3fdc9278a40e22efbee7","last_reissued_at":"2026-07-14T02:22:17.584461Z","signature_status":"signed_v1","first_computed_at":"2026-07-14T02:22:17.584461Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How to Tame Grokking: Representation Geometry as a Control Signal","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Maksim A Kazanskii","submitted_at":"2026-07-13T15:09:42Z","abstract_excerpt":"Grokking is a phenomenon in which neural networks initially memorize training data and only later exhibit strong generalization after prolonged optimization. Despite extensive recent study, the factors influencing the emergence and timing of grokking remain incompletely understood. We investigate the relationship between representation geometry and delayed generalization. We find that dimensionality collapse consistently precedes the onset of grokking in all evaluated settings. Motivated by these observations, we introduce Geometric Dimensionality Regularization (GeomDR), a simple spectral reg"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.11666","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/2607.11666/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":"2607.11666","created_at":"2026-07-14T02:22:17.584915+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.11666v1","created_at":"2026-07-14T02:22:17.584915+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.11666","created_at":"2026-07-14T02:22:17.584915+00:00"},{"alias_kind":"pith_short_12","alias_value":"JQX2LNR4MW6D","created_at":"2026-07-14T02:22:17.584915+00:00"},{"alias_kind":"pith_short_16","alias_value":"JQX2LNR4MW6DTPSZ","created_at":"2026-07-14T02:22:17.584915+00:00"},{"alias_kind":"pith_short_8","alias_value":"JQX2LNR4","created_at":"2026-07-14T02:22:17.584915+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS","json":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS.json","graph_json":"https://pith.science/api/pith-number/JQX2LNR4MW6DTPSZI46444CLAS/graph.json","events_json":"https://pith.science/api/pith-number/JQX2LNR4MW6DTPSZI46444CLAS/events.json","paper":"https://pith.science/paper/JQX2LNR4"},"agent_actions":{"view_html":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS","download_json":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS.json","view_paper":"https://pith.science/paper/JQX2LNR4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.11666&json=true","fetch_graph":"https://pith.science/api/pith-number/JQX2LNR4MW6DTPSZI46444CLAS/graph.json","fetch_events":"https://pith.science/api/pith-number/JQX2LNR4MW6DTPSZI46444CLAS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS/action/storage_attestation","attest_author":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS/action/author_attestation","sign_citation":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS/action/citation_signature","submit_replication":"https://pith.science/pith/JQX2LNR4MW6DTPSZI46444CLAS/action/replication_record"}},"created_at":"2026-07-14T02:22:17.584915+00:00","updated_at":"2026-07-14T02:22:17.584915+00:00"}