{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:PRCOOHU6MKQGCAY3OUPTKCIOKZ","short_pith_number":"pith:PRCOOHU6","schema_version":"1.0","canonical_sha256":"7c44e71e9e62a061031b751f35090e567e6cdb49a6c93af73b4d167d80b0a389","source":{"kind":"arxiv","id":"2006.00719","version":3},"attestation_state":"computed","paper":{"title":"ADAHESSIAN: An Adaptive Second Order Optimizer for Machine Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA","math.NA","stat.ML"],"primary_cat":"cs.LG","authors_text":"Amir Gholami, Kurt Keutzer, Michael W. Mahoney, Mustafa Mustafa, Sheng Shen, Zhewei Yao","submitted_at":"2020-06-01T05:00:51Z","abstract_excerpt":"We introduce ADAHESSIAN, a second order stochastic optimization algorithm which dynamically incorporates the curvature of the loss function via ADAptive estimates of the HESSIAN. Second order algorithms are among the most powerful optimization algorithms with superior convergence properties as compared to first order methods such as SGD and Adam. The main disadvantage of traditional second order methods is their heavier per iteration computation and poor accuracy as compared to first order methods. To address these, we incorporate several novel approaches in ADAHESSIAN, including: (i) a fast H"},"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":"2006.00719","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-01T05:00:51Z","cross_cats_sorted":["cs.NA","math.NA","stat.ML"],"title_canon_sha256":"cc447e6ca4c771fe9fb0bf88670fa5cb21ca1ebc5f30331b77bef73816cd0c07","abstract_canon_sha256":"69c4418c31a98b999010fa602935aae7c6a8bfe3acdf6f92a44af467c762cf30"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:36:07.412592Z","signature_b64":"09czXraPxyz0O4t4xGeWf2Uh7avxOwY9/IvHu2nkIm4OpA1OTZ8dM66OSP4rA0kthXdZIIkgB/nBRVK6kWSJDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7c44e71e9e62a061031b751f35090e567e6cdb49a6c93af73b4d167d80b0a389","last_reissued_at":"2026-07-05T02:36:07.412047Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:36:07.412047Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ADAHESSIAN: An Adaptive Second Order Optimizer for Machine Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA","math.NA","stat.ML"],"primary_cat":"cs.LG","authors_text":"Amir Gholami, Kurt Keutzer, Michael W. Mahoney, Mustafa Mustafa, Sheng Shen, Zhewei Yao","submitted_at":"2020-06-01T05:00:51Z","abstract_excerpt":"We introduce ADAHESSIAN, a second order stochastic optimization algorithm which dynamically incorporates the curvature of the loss function via ADAptive estimates of the HESSIAN. Second order algorithms are among the most powerful optimization algorithms with superior convergence properties as compared to first order methods such as SGD and Adam. The main disadvantage of traditional second order methods is their heavier per iteration computation and poor accuracy as compared to first order methods. To address these, we incorporate several novel approaches in ADAHESSIAN, including: (i) a fast H"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2006.00719","kind":"arxiv","version":3},"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/2006.00719/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":"2006.00719","created_at":"2026-07-05T02:36:07.412139+00:00"},{"alias_kind":"arxiv_version","alias_value":"2006.00719v3","created_at":"2026-07-05T02:36:07.412139+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.00719","created_at":"2026-07-05T02:36:07.412139+00:00"},{"alias_kind":"pith_short_12","alias_value":"PRCOOHU6MKQG","created_at":"2026-07-05T02:36:07.412139+00:00"},{"alias_kind":"pith_short_16","alias_value":"PRCOOHU6MKQGCAY3","created_at":"2026-07-05T02:36:07.412139+00:00"},{"alias_kind":"pith_short_8","alias_value":"PRCOOHU6","created_at":"2026-07-05T02:36:07.412139+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2504.17618","citing_title":"The effects of Hessian eigenvalue spectral density type on the applicability of Hessian analysis to generalization capability assessment of neural networks","ref_index":19,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ","json":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ.json","graph_json":"https://pith.science/api/pith-number/PRCOOHU6MKQGCAY3OUPTKCIOKZ/graph.json","events_json":"https://pith.science/api/pith-number/PRCOOHU6MKQGCAY3OUPTKCIOKZ/events.json","paper":"https://pith.science/paper/PRCOOHU6"},"agent_actions":{"view_html":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ","download_json":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ.json","view_paper":"https://pith.science/paper/PRCOOHU6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2006.00719&json=true","fetch_graph":"https://pith.science/api/pith-number/PRCOOHU6MKQGCAY3OUPTKCIOKZ/graph.json","fetch_events":"https://pith.science/api/pith-number/PRCOOHU6MKQGCAY3OUPTKCIOKZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ/action/storage_attestation","attest_author":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ/action/author_attestation","sign_citation":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ/action/citation_signature","submit_replication":"https://pith.science/pith/PRCOOHU6MKQGCAY3OUPTKCIOKZ/action/replication_record"}},"created_at":"2026-07-05T02:36:07.412139+00:00","updated_at":"2026-07-05T02:36:07.412139+00:00"}