{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:VAS5YOWAXVNAXW7XIUITAYDZIP","short_pith_number":"pith:VAS5YOWA","schema_version":"1.0","canonical_sha256":"a825dc3ac0bd5a0bdbf7451130607943f536ca4120271f9c655a894573d46ff9","source":{"kind":"arxiv","id":"2103.01931","version":2},"attestation_state":"computed","paper":{"title":"Categorical Foundations of Gradient-Based Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.CT"],"primary_cat":"cs.LG","authors_text":"Bruno Gavranovi\\'c, Fabio Zanasi, G.S.H. Cruttwell, Neil Ghani, Paul Wilson","submitted_at":"2021-03-02T18:43:10Z","abstract_excerpt":"We propose a categorical semantics of gradient-based machine learning algorithms in terms of lenses, parametrised maps, and reverse derivative categories. This foundation provides a powerful explanatory and unifying framework: it encompasses a variety of gradient descent algorithms such as ADAM, AdaGrad, and Nesterov momentum, as well as a variety of loss functions such as as MSE and Softmax cross-entropy, shedding new light on their similarities and differences. Our approach to gradient-based learning has examples generalising beyond the familiar continuous domains (modelled in categories of "},"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":"2103.01931","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-03-02T18:43:10Z","cross_cats_sorted":["math.CT"],"title_canon_sha256":"37ba30f58e2a7c17edf14ccf554d621020c4d6eb8898e06bcc62cb0648f770a5","abstract_canon_sha256":"8b57afbda9a9a5aabf32132d75debac4ed48564d0a0fd0e18a7070d4c1bd2d0b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:57:16.811697Z","signature_b64":"9SWjdTeSGdGRvYZLSXJetBVU3sGa2z2UsiHgNZ8h5ql4ZXy0bWpg4T9GaKthRg5m+ANL/BvT7HpdnDLfP7K3Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a825dc3ac0bd5a0bdbf7451130607943f536ca4120271f9c655a894573d46ff9","last_reissued_at":"2026-07-05T02:57:16.811202Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:57:16.811202Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Categorical Foundations of Gradient-Based Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.CT"],"primary_cat":"cs.LG","authors_text":"Bruno Gavranovi\\'c, Fabio Zanasi, G.S.H. Cruttwell, Neil Ghani, Paul Wilson","submitted_at":"2021-03-02T18:43:10Z","abstract_excerpt":"We propose a categorical semantics of gradient-based machine learning algorithms in terms of lenses, parametrised maps, and reverse derivative categories. This foundation provides a powerful explanatory and unifying framework: it encompasses a variety of gradient descent algorithms such as ADAM, AdaGrad, and Nesterov momentum, as well as a variety of loss functions such as as MSE and Softmax cross-entropy, shedding new light on their similarities and differences. Our approach to gradient-based learning has examples generalising beyond the familiar continuous domains (modelled in categories of "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2103.01931","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/2103.01931/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":"2103.01931","created_at":"2026-07-05T02:57:16.811264+00:00"},{"alias_kind":"arxiv_version","alias_value":"2103.01931v2","created_at":"2026-07-05T02:57:16.811264+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2103.01931","created_at":"2026-07-05T02:57:16.811264+00:00"},{"alias_kind":"pith_short_12","alias_value":"VAS5YOWAXVNA","created_at":"2026-07-05T02:57:16.811264+00:00"},{"alias_kind":"pith_short_16","alias_value":"VAS5YOWAXVNAXW7X","created_at":"2026-07-05T02:57:16.811264+00:00"},{"alias_kind":"pith_short_8","alias_value":"VAS5YOWA","created_at":"2026-07-05T02:57:16.811264+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2207.09180","citing_title":"Polycategorical Constructions for Unitary Supermaps of Arbitrary Dimension","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP","json":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP.json","graph_json":"https://pith.science/api/pith-number/VAS5YOWAXVNAXW7XIUITAYDZIP/graph.json","events_json":"https://pith.science/api/pith-number/VAS5YOWAXVNAXW7XIUITAYDZIP/events.json","paper":"https://pith.science/paper/VAS5YOWA"},"agent_actions":{"view_html":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP","download_json":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP.json","view_paper":"https://pith.science/paper/VAS5YOWA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2103.01931&json=true","fetch_graph":"https://pith.science/api/pith-number/VAS5YOWAXVNAXW7XIUITAYDZIP/graph.json","fetch_events":"https://pith.science/api/pith-number/VAS5YOWAXVNAXW7XIUITAYDZIP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP/action/storage_attestation","attest_author":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP/action/author_attestation","sign_citation":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP/action/citation_signature","submit_replication":"https://pith.science/pith/VAS5YOWAXVNAXW7XIUITAYDZIP/action/replication_record"}},"created_at":"2026-07-05T02:57:16.811264+00:00","updated_at":"2026-07-05T02:57:16.811264+00:00"}