{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SNBONQASJJG67LIGUUIZJ6HOGD","short_pith_number":"pith:SNBONQAS","schema_version":"1.0","canonical_sha256":"9342e6c0124a4defad06a51194f8ee30e2ce387f151000844692bd9ebec2af26","source":{"kind":"arxiv","id":"2606.28662","version":1},"attestation_state":"computed","paper":{"title":"Closed-Form Steepest Descent Direction toward Flat Minima: Reducing Upper Bounds on the Loss Hessian Eigenspectrum in Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.LG","authors_text":"Hirotaka Takahashi, Kazuki Sakai, Makoto Sasaki, Yohei Kakimoto, Yusuke Sakai, Yuto Omae","submitted_at":"2026-06-27T00:47:03Z","abstract_excerpt":"The flatness hypothesis suggests that flatness of the loss landscape, as measured by the eigenvalues of the loss Hessian, correlates with better neural network generalization. While various algorithms reduce these eigenvalues, most focus on procedural design, leaving it unclear how data distributions and NN parameters structurally determine directions toward flat minima. Characterizing these directions analytically is generally intractable. To overcome this mathematical difficulty, recent studies derived the Wolkowicz-Styan (WS) upper bound on the maximum eigenvalue of the cross-entropy loss 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":"2606.28662","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-06-27T00:47:03Z","cross_cats_sorted":["cs.AI","cs.NE"],"title_canon_sha256":"4e9b4528858c812110de0b184a54aa9f5c836753be75403307ed1872cec0f3d2","abstract_canon_sha256":"b9376bbcb10f8bd410fb3dd17c25bd9a17adbac14bce24643a4eca76b6aed836"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:16:46.419457Z","signature_b64":"lrueJZAleI7kGKX81h5AL39qeElPPgykzeqRQNGnfpotjIn0YxvjTolhf8k2DPFGuOVtrgQ/5FloraWwaN0SCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9342e6c0124a4defad06a51194f8ee30e2ce387f151000844692bd9ebec2af26","last_reissued_at":"2026-06-30T01:16:46.418554Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:16:46.418554Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Closed-Form Steepest Descent Direction toward Flat Minima: Reducing Upper Bounds on the Loss Hessian Eigenspectrum in Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.LG","authors_text":"Hirotaka Takahashi, Kazuki Sakai, Makoto Sasaki, Yohei Kakimoto, Yusuke Sakai, Yuto Omae","submitted_at":"2026-06-27T00:47:03Z","abstract_excerpt":"The flatness hypothesis suggests that flatness of the loss landscape, as measured by the eigenvalues of the loss Hessian, correlates with better neural network generalization. While various algorithms reduce these eigenvalues, most focus on procedural design, leaving it unclear how data distributions and NN parameters structurally determine directions toward flat minima. Characterizing these directions analytically is generally intractable. To overcome this mathematical difficulty, recent studies derived the Wolkowicz-Styan (WS) upper bound on the maximum eigenvalue of the cross-entropy loss H"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.28662","kind":"arxiv","version":1},"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/2606.28662/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":"2606.28662","created_at":"2026-06-30T01:16:46.418685+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.28662v1","created_at":"2026-06-30T01:16:46.418685+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.28662","created_at":"2026-06-30T01:16:46.418685+00:00"},{"alias_kind":"pith_short_12","alias_value":"SNBONQASJJG6","created_at":"2026-06-30T01:16:46.418685+00:00"},{"alias_kind":"pith_short_16","alias_value":"SNBONQASJJG67LIG","created_at":"2026-06-30T01:16:46.418685+00:00"},{"alias_kind":"pith_short_8","alias_value":"SNBONQAS","created_at":"2026-06-30T01:16:46.418685+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/SNBONQASJJG67LIGUUIZJ6HOGD","json":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD.json","graph_json":"https://pith.science/api/pith-number/SNBONQASJJG67LIGUUIZJ6HOGD/graph.json","events_json":"https://pith.science/api/pith-number/SNBONQASJJG67LIGUUIZJ6HOGD/events.json","paper":"https://pith.science/paper/SNBONQAS"},"agent_actions":{"view_html":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD","download_json":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD.json","view_paper":"https://pith.science/paper/SNBONQAS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.28662&json=true","fetch_graph":"https://pith.science/api/pith-number/SNBONQASJJG67LIGUUIZJ6HOGD/graph.json","fetch_events":"https://pith.science/api/pith-number/SNBONQASJJG67LIGUUIZJ6HOGD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD/action/storage_attestation","attest_author":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD/action/author_attestation","sign_citation":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD/action/citation_signature","submit_replication":"https://pith.science/pith/SNBONQASJJG67LIGUUIZJ6HOGD/action/replication_record"}},"created_at":"2026-06-30T01:16:46.418685+00:00","updated_at":"2026-06-30T01:16:46.418685+00:00"}