{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:LL6CSBDLGGX7ELCPO4Q5DY3X3H","short_pith_number":"pith:LL6CSBDL","schema_version":"1.0","canonical_sha256":"5afc29046b31aff22c4f7721d1e377d9f73a4616dd2836a229f20ca51f4604ff","source":{"kind":"arxiv","id":"2305.17332","version":2},"attestation_state":"computed","paper":{"title":"Learning Capacity: A Measure of the Effective Dimensionality of a Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","math.IT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Daiwei Chen, Pratik Chaudhari, Wei-Kai Chang","submitted_at":"2023-05-27T02:27:27Z","abstract_excerpt":"We use a formal correspondence between thermodynamics and inference, where the number of samples can be thought of as the inverse temperature, to study a quantity called ``learning capacity'' which is a measure of the effective dimensionality of a model. We show that the learning capacity is a useful notion of the complexity because (a) it correlates well with the test loss and it is a tiny fraction of the number of parameters for many deep networks trained on typical datasets, (b) it depends upon the number of samples used for training, (c) it is numerically consistent with notions of capacit"},"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":"2305.17332","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-05-27T02:27:27Z","cross_cats_sorted":["cs.IT","math.IT","stat.ML"],"title_canon_sha256":"8f8bb7fd759e241de170feea9a55c8d08c3e38616d1ada7fd47a07c0bd725345","abstract_canon_sha256":"5804ed4249d4a8675d7a0a2165b5c5a3c22320843716c22bec20043d26b78862"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:22:59.372936Z","signature_b64":"mJemIc5aV0Y5v7KHAfpWtfOjBsAOaS/w7Ew/CNLVY64fw1nil2k7NQei7utOm5+7sYC73IFO7p23oEPIbNSwBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5afc29046b31aff22c4f7721d1e377d9f73a4616dd2836a229f20ca51f4604ff","last_reissued_at":"2026-07-05T09:22:59.372440Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:22:59.372440Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Capacity: A Measure of the Effective Dimensionality of a Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","math.IT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Daiwei Chen, Pratik Chaudhari, Wei-Kai Chang","submitted_at":"2023-05-27T02:27:27Z","abstract_excerpt":"We use a formal correspondence between thermodynamics and inference, where the number of samples can be thought of as the inverse temperature, to study a quantity called ``learning capacity'' which is a measure of the effective dimensionality of a model. We show that the learning capacity is a useful notion of the complexity because (a) it correlates well with the test loss and it is a tiny fraction of the number of parameters for many deep networks trained on typical datasets, (b) it depends upon the number of samples used for training, (c) it is numerically consistent with notions of capacit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.17332","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/2305.17332/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":"2305.17332","created_at":"2026-07-05T09:22:59.372495+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.17332v2","created_at":"2026-07-05T09:22:59.372495+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.17332","created_at":"2026-07-05T09:22:59.372495+00:00"},{"alias_kind":"pith_short_12","alias_value":"LL6CSBDLGGX7","created_at":"2026-07-05T09:22:59.372495+00:00"},{"alias_kind":"pith_short_16","alias_value":"LL6CSBDLGGX7ELCP","created_at":"2026-07-05T09:22:59.372495+00:00"},{"alias_kind":"pith_short_8","alias_value":"LL6CSBDL","created_at":"2026-07-05T09:22:59.372495+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.24477","citing_title":"The Normalized Maximum Likelihood for Regular Non-Smooth Models: Measure-Theoretic Foundations and Geometric Sampling","ref_index":38,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H","json":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H.json","graph_json":"https://pith.science/api/pith-number/LL6CSBDLGGX7ELCPO4Q5DY3X3H/graph.json","events_json":"https://pith.science/api/pith-number/LL6CSBDLGGX7ELCPO4Q5DY3X3H/events.json","paper":"https://pith.science/paper/LL6CSBDL"},"agent_actions":{"view_html":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H","download_json":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H.json","view_paper":"https://pith.science/paper/LL6CSBDL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.17332&json=true","fetch_graph":"https://pith.science/api/pith-number/LL6CSBDLGGX7ELCPO4Q5DY3X3H/graph.json","fetch_events":"https://pith.science/api/pith-number/LL6CSBDLGGX7ELCPO4Q5DY3X3H/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H/action/storage_attestation","attest_author":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H/action/author_attestation","sign_citation":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H/action/citation_signature","submit_replication":"https://pith.science/pith/LL6CSBDLGGX7ELCPO4Q5DY3X3H/action/replication_record"}},"created_at":"2026-07-05T09:22:59.372495+00:00","updated_at":"2026-07-05T09:22:59.372495+00:00"}