{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:KV6JDHIJC2PAUQRT5AJJS6LQL6","short_pith_number":"pith:KV6JDHIJ","schema_version":"1.0","canonical_sha256":"557c919d09169e0a4233e8129979705fa84ef8d8a1c18db451d2e8065b5e9bbc","source":{"kind":"arxiv","id":"1806.03884","version":2},"attestation_state":"computed","paper":{"title":"Fast Approximate Natural Gradient Descent in a Kronecker-factored Eigenbasis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"C\\'esar Laurent, Nicolas Ballas, Pascal Vincent, Thomas George, Xavier Bouthillier","submitted_at":"2018-06-11T09:44:23Z","abstract_excerpt":"Optimization algorithms that leverage gradient covariance information, such as variants of natural gradient descent (Amari, 1998), offer the prospect of yielding more effective descent directions. For models with many parameters, the covariance matrix they are based on becomes gigantic, making them inapplicable in their original form. This has motivated research into both simple diagonal approximations and more sophisticated factored approximations such as KFAC (Heskes, 2000; Martens & Grosse, 2015; Grosse & Martens, 2016). In the present work we draw inspiration from both to propose a novel a"},"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":"1806.03884","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-06-11T09:44:23Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"48018c919b891584e8e0a2ecae90e6e0100d84de69305b15b60c9acfa9f9b76c","abstract_canon_sha256":"7ad2ba36e9d0bda8b32cd1fd26307498d300c0e26eefb77232c1858f2e96e43f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:00:25.928536Z","signature_b64":"afOGswdiYGpb/G668U5E2DYKhsT4TOKtmDuYQur7JD4zl4jg11VYoXsAHCw5Z8JUUyLblGnY1PVxXxN5vJI7DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"557c919d09169e0a4233e8129979705fa84ef8d8a1c18db451d2e8065b5e9bbc","last_reissued_at":"2026-07-05T03:00:25.928123Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:00:25.928123Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fast Approximate Natural Gradient Descent in a Kronecker-factored Eigenbasis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"C\\'esar Laurent, Nicolas Ballas, Pascal Vincent, Thomas George, Xavier Bouthillier","submitted_at":"2018-06-11T09:44:23Z","abstract_excerpt":"Optimization algorithms that leverage gradient covariance information, such as variants of natural gradient descent (Amari, 1998), offer the prospect of yielding more effective descent directions. For models with many parameters, the covariance matrix they are based on becomes gigantic, making them inapplicable in their original form. This has motivated research into both simple diagonal approximations and more sophisticated factored approximations such as KFAC (Heskes, 2000; Martens & Grosse, 2015; Grosse & Martens, 2016). In the present work we draw inspiration from both to propose a novel a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1806.03884","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/1806.03884/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":"1806.03884","created_at":"2026-07-05T03:00:25.928184+00:00"},{"alias_kind":"arxiv_version","alias_value":"1806.03884v2","created_at":"2026-07-05T03:00:25.928184+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1806.03884","created_at":"2026-07-05T03:00:25.928184+00:00"},{"alias_kind":"pith_short_12","alias_value":"KV6JDHIJC2PA","created_at":"2026-07-05T03:00:25.928184+00:00"},{"alias_kind":"pith_short_16","alias_value":"KV6JDHIJC2PAUQRT","created_at":"2026-07-05T03:00:25.928184+00:00"},{"alias_kind":"pith_short_8","alias_value":"KV6JDHIJ","created_at":"2026-07-05T03:00:25.928184+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07204","citing_title":"Restricted Dynamic Geometric Complexity: Certificates for Structured Preconditioning","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2606.30813","citing_title":"Gradient Smoothing: Coupling Layer-wise Updates for Improved Optimization","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6","json":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6.json","graph_json":"https://pith.science/api/pith-number/KV6JDHIJC2PAUQRT5AJJS6LQL6/graph.json","events_json":"https://pith.science/api/pith-number/KV6JDHIJC2PAUQRT5AJJS6LQL6/events.json","paper":"https://pith.science/paper/KV6JDHIJ"},"agent_actions":{"view_html":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6","download_json":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6.json","view_paper":"https://pith.science/paper/KV6JDHIJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1806.03884&json=true","fetch_graph":"https://pith.science/api/pith-number/KV6JDHIJC2PAUQRT5AJJS6LQL6/graph.json","fetch_events":"https://pith.science/api/pith-number/KV6JDHIJC2PAUQRT5AJJS6LQL6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6/action/storage_attestation","attest_author":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6/action/author_attestation","sign_citation":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6/action/citation_signature","submit_replication":"https://pith.science/pith/KV6JDHIJC2PAUQRT5AJJS6LQL6/action/replication_record"}},"created_at":"2026-07-05T03:00:25.928184+00:00","updated_at":"2026-07-05T03:00:25.928184+00:00"}