{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:5HSEEXCWAB2GX6XHEO27ABWGHB","short_pith_number":"pith:5HSEEXCW","schema_version":"1.0","canonical_sha256":"e9e4425c5600746bfae723b5f006c638433b7798352e7b2837ac377bf69d62ae","source":{"kind":"arxiv","id":"2507.20057","version":1},"attestation_state":"computed","paper":{"title":"What Can Grokking Teach Us About Learning Under Nonstationarity?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andras Gyorgy, Clare Lyle, Gharda Sokar, Razvan Pascanu","submitted_at":"2025-07-26T20:51:24Z","abstract_excerpt":"In continual learning problems, it is often necessary to overwrite components of a neural network's learned representation in response to changes in the data stream; however, neural networks often exhibit \\primacy bias, whereby early training data hinders the network's ability to generalize on later tasks. While feature-learning dynamics of nonstationary learning problems are not well studied, the emergence of feature-learning dynamics is known to drive the phenomenon of grokking, wherein neural networks initially memorize their training data and only later exhibit perfect generalization. This"},"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":"2507.20057","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-26T20:51:24Z","cross_cats_sorted":[],"title_canon_sha256":"9c3922d8b33fb8e6311e7424ec993a30df56ccec5df25c3115bcd2fd2a368fd4","abstract_canon_sha256":"f56df9ea69195ee8fa0e6dc7453620933d70b9af09695a5621da4d4e038a3230"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:12.631524Z","signature_b64":"uUpBmSpEHLKjJ8fQFxwrPVh6cofDiRZMxjPXWm6MBu9zGQOUrMNiBzQBjlMhU4yP0kleEOvsuB4erwW6NkIBAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e9e4425c5600746bfae723b5f006c638433b7798352e7b2837ac377bf69d62ae","last_reissued_at":"2026-07-05T11:44:12.631072Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:12.631072Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What Can Grokking Teach Us About Learning Under Nonstationarity?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andras Gyorgy, Clare Lyle, Gharda Sokar, Razvan Pascanu","submitted_at":"2025-07-26T20:51:24Z","abstract_excerpt":"In continual learning problems, it is often necessary to overwrite components of a neural network's learned representation in response to changes in the data stream; however, neural networks often exhibit \\primacy bias, whereby early training data hinders the network's ability to generalize on later tasks. While feature-learning dynamics of nonstationary learning problems are not well studied, the emergence of feature-learning dynamics is known to drive the phenomenon of grokking, wherein neural networks initially memorize their training data and only later exhibit perfect generalization. This"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.20057","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/2507.20057/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":"2507.20057","created_at":"2026-07-05T11:44:12.631135+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.20057v1","created_at":"2026-07-05T11:44:12.631135+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.20057","created_at":"2026-07-05T11:44:12.631135+00:00"},{"alias_kind":"pith_short_12","alias_value":"5HSEEXCWAB2G","created_at":"2026-07-05T11:44:12.631135+00:00"},{"alias_kind":"pith_short_16","alias_value":"5HSEEXCWAB2GX6XH","created_at":"2026-07-05T11:44:12.631135+00:00"},{"alias_kind":"pith_short_8","alias_value":"5HSEEXCW","created_at":"2026-07-05T11:44:12.631135+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20441","citing_title":"Weight Decay Regimes in Grokking Transformers: Cheap Online Diagnostics","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB","json":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB.json","graph_json":"https://pith.science/api/pith-number/5HSEEXCWAB2GX6XHEO27ABWGHB/graph.json","events_json":"https://pith.science/api/pith-number/5HSEEXCWAB2GX6XHEO27ABWGHB/events.json","paper":"https://pith.science/paper/5HSEEXCW"},"agent_actions":{"view_html":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB","download_json":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB.json","view_paper":"https://pith.science/paper/5HSEEXCW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.20057&json=true","fetch_graph":"https://pith.science/api/pith-number/5HSEEXCWAB2GX6XHEO27ABWGHB/graph.json","fetch_events":"https://pith.science/api/pith-number/5HSEEXCWAB2GX6XHEO27ABWGHB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB/action/storage_attestation","attest_author":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB/action/author_attestation","sign_citation":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB/action/citation_signature","submit_replication":"https://pith.science/pith/5HSEEXCWAB2GX6XHEO27ABWGHB/action/replication_record"}},"created_at":"2026-07-05T11:44:12.631135+00:00","updated_at":"2026-07-05T11:44:12.631135+00:00"}