{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:P2RV67TXEODHJAOUC6SRYUQW55","short_pith_number":"pith:P2RV67TX","schema_version":"1.0","canonical_sha256":"7ea35f7e7723867481d417a51c5216ef515df5134f2f2fbd1dbe679c6067e085","source":{"kind":"arxiv","id":"2111.05478","version":2},"attestation_state":"computed","paper":{"title":"SGD Through the Lens of Kolmogorov Complexity","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Gregory Schwartzman","submitted_at":"2021-11-10T01:32:38Z","abstract_excerpt":"We prove that stochastic gradient descent (SGD) finds a solution that achieves $(1-\\epsilon)$ classification accuracy on the entire dataset. We do so under two main assumptions: (1. Local progress) The model accuracy improves on average over batches. (2. Models compute simple functions) The function computed by the model is simple (has low Kolmogorov complexity). It is sufficient that these assumptions hold only for a tiny fraction of the epochs.\n  Intuitively, the above means that intermittent local progress of SGD implies global progress. Assumption 2 trivially holds for underparameterized m"},"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":"2111.05478","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-11-10T01:32:38Z","cross_cats_sorted":[],"title_canon_sha256":"56606588087e882dc57dfb0b2eb9083b2ad534833fb422fe3c9c70d7348752c8","abstract_canon_sha256":"9521c26760951d5dbf3b1c153e819aed19f0859c0ca85ecf5dd81b3a88f2a983"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:23:05.615697Z","signature_b64":"iQJFIegEm2BTKDjSzC3bRyekjKEloaYksc3pKjT/vMI17GHu7BZMnElb9LK+khVJzG6BhASFVFB1LmCvt/fpAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7ea35f7e7723867481d417a51c5216ef515df5134f2f2fbd1dbe679c6067e085","last_reissued_at":"2026-07-05T04:23:05.615275Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:23:05.615275Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SGD Through the Lens of Kolmogorov Complexity","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Gregory Schwartzman","submitted_at":"2021-11-10T01:32:38Z","abstract_excerpt":"We prove that stochastic gradient descent (SGD) finds a solution that achieves $(1-\\epsilon)$ classification accuracy on the entire dataset. We do so under two main assumptions: (1. Local progress) The model accuracy improves on average over batches. (2. Models compute simple functions) The function computed by the model is simple (has low Kolmogorov complexity). It is sufficient that these assumptions hold only for a tiny fraction of the epochs.\n  Intuitively, the above means that intermittent local progress of SGD implies global progress. Assumption 2 trivially holds for underparameterized m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.05478","kind":"arxiv","version":2},"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/2111.05478/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":"2111.05478","created_at":"2026-07-05T04:23:05.615342+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.05478v2","created_at":"2026-07-05T04:23:05.615342+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.05478","created_at":"2026-07-05T04:23:05.615342+00:00"},{"alias_kind":"pith_short_12","alias_value":"P2RV67TXEODH","created_at":"2026-07-05T04:23:05.615342+00:00"},{"alias_kind":"pith_short_16","alias_value":"P2RV67TXEODHJAOU","created_at":"2026-07-05T04:23:05.615342+00:00"},{"alias_kind":"pith_short_8","alias_value":"P2RV67TX","created_at":"2026-07-05T04:23:05.615342+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/P2RV67TXEODHJAOUC6SRYUQW55","json":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55.json","graph_json":"https://pith.science/api/pith-number/P2RV67TXEODHJAOUC6SRYUQW55/graph.json","events_json":"https://pith.science/api/pith-number/P2RV67TXEODHJAOUC6SRYUQW55/events.json","paper":"https://pith.science/paper/P2RV67TX"},"agent_actions":{"view_html":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55","download_json":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55.json","view_paper":"https://pith.science/paper/P2RV67TX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.05478&json=true","fetch_graph":"https://pith.science/api/pith-number/P2RV67TXEODHJAOUC6SRYUQW55/graph.json","fetch_events":"https://pith.science/api/pith-number/P2RV67TXEODHJAOUC6SRYUQW55/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55/action/storage_attestation","attest_author":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55/action/author_attestation","sign_citation":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55/action/citation_signature","submit_replication":"https://pith.science/pith/P2RV67TXEODHJAOUC6SRYUQW55/action/replication_record"}},"created_at":"2026-07-05T04:23:05.615342+00:00","updated_at":"2026-07-05T04:23:05.615342+00:00"}