{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:7FSFJDXKYX337CB2D5CTDZBIR5","short_pith_number":"pith:7FSFJDXK","canonical_record":{"source":{"id":"2502.06268","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"stat.ML","submitted_at":"2025-02-10T09:07:04Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"10e171fd4d5e3aa5788b0d82b18bac139c7e018eaed62f59d7a15c715d5c8f83","abstract_canon_sha256":"87047074fc315447bc29e9ef0202f65c358704ea42a608e28f0c3e4f0b335054"},"schema_version":"1.0"},"canonical_sha256":"f964548eeac5f7bf883a1f4531e4288f4052e9da7b49fde0b94135b24c84e5d2","source":{"kind":"arxiv","id":"2502.06268","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06268","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06268v3","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06268","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"pith_short_12","alias_value":"7FSFJDXKYX33","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"pith_short_16","alias_value":"7FSFJDXKYX337CB2","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"pith_short_8","alias_value":"7FSFJDXK","created_at":"2026-07-05T10:40:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:7FSFJDXKYX337CB2D5CTDZBIR5","target":"record","payload":{"canonical_record":{"source":{"id":"2502.06268","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"stat.ML","submitted_at":"2025-02-10T09:07:04Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"10e171fd4d5e3aa5788b0d82b18bac139c7e018eaed62f59d7a15c715d5c8f83","abstract_canon_sha256":"87047074fc315447bc29e9ef0202f65c358704ea42a608e28f0c3e4f0b335054"},"schema_version":"1.0"},"canonical_sha256":"f964548eeac5f7bf883a1f4531e4288f4052e9da7b49fde0b94135b24c84e5d2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:40:59.897412Z","signature_b64":"5MFlidflo8sgBxlYljLSSK+2ULXYl5Q+biK/jSNY+6WXUaKEPMkXmkhUH3k7LgtJFGji47AJdNWJ2l9JbXCGBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f964548eeac5f7bf883a1f4531e4288f4052e9da7b49fde0b94135b24c84e5d2","last_reissued_at":"2026-07-05T10:40:59.896903Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:40:59.896903Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.06268","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:40:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JW2YGUTqQYBbth0XTqnDiA8+4TglfXXagJvq7qB0TNlfPD0kBd9dis8axMlhvscjhpZpDtLhxRmxxROEBa9DAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T11:36:32.877291Z"},"content_sha256":"b2f5dd9771169f9092cbfe69574442eb20e2fa7dba13be28d7be1e732cdfa1fb","schema_version":"1.0","event_id":"sha256:b2f5dd9771169f9092cbfe69574442eb20e2fa7dba13be28d7be1e732cdfa1fb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:7FSFJDXKYX337CB2D5CTDZBIR5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Spectral-factorized Positive-definite Curvature Learning for NN Training","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Felix Dangel, Juhan Bae, Richard E. Turner, Roger B. Grosse, Runa Eschenhagen, Wu Lin","submitted_at":"2025-02-10T09:07:04Z","abstract_excerpt":"Many training methods, such as Adam(W) and Shampoo, learn a positive-definite curvature matrix and apply an inverse root before preconditioning. Recently, non-diagonal training methods, such as Shampoo, have gained significant attention; however, they remain computationally inefficient and are limited to specific types of curvature information due to the costly matrix root computation via matrix decomposition. To address this, we propose a Riemannian optimization approach that dynamically adapts spectral-factorized positive-definite curvature estimates, enabling the efficient application of ar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06268","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/2502.06268/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:40:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E3FGV7k9Qv0FAZxHJ1rCZlrMG+Dkt5+I9E0RWvb0D7ZlBxVJmXn1aJfkUYZmnrQPQZsC2frArcYC7LUuNw05CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T11:36:32.878342Z"},"content_sha256":"bcfdf40713029623753ac514666831410e54c90e86e83508294f4604ec6ab428","schema_version":"1.0","event_id":"sha256:bcfdf40713029623753ac514666831410e54c90e86e83508294f4604ec6ab428"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7FSFJDXKYX337CB2D5CTDZBIR5/bundle.json","state_url":"https://pith.science/pith/7FSFJDXKYX337CB2D5CTDZBIR5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7FSFJDXKYX337CB2D5CTDZBIR5/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-09T11:36:32Z","links":{"resolver":"https://pith.science/pith/7FSFJDXKYX337CB2D5CTDZBIR5","bundle":"https://pith.science/pith/7FSFJDXKYX337CB2D5CTDZBIR5/bundle.json","state":"https://pith.science/pith/7FSFJDXKYX337CB2D5CTDZBIR5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7FSFJDXKYX337CB2D5CTDZBIR5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:7FSFJDXKYX337CB2D5CTDZBIR5","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"87047074fc315447bc29e9ef0202f65c358704ea42a608e28f0c3e4f0b335054","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"stat.ML","submitted_at":"2025-02-10T09:07:04Z","title_canon_sha256":"10e171fd4d5e3aa5788b0d82b18bac139c7e018eaed62f59d7a15c715d5c8f83"},"schema_version":"1.0","source":{"id":"2502.06268","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06268","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06268v3","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06268","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"pith_short_12","alias_value":"7FSFJDXKYX33","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"pith_short_16","alias_value":"7FSFJDXKYX337CB2","created_at":"2026-07-05T10:40:59Z"},{"alias_kind":"pith_short_8","alias_value":"7FSFJDXK","created_at":"2026-07-05T10:40:59Z"}],"graph_snapshots":[{"event_id":"sha256:bcfdf40713029623753ac514666831410e54c90e86e83508294f4604ec6ab428","target":"graph","created_at":"2026-07-05T10:40:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.06268/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Many training methods, such as Adam(W) and Shampoo, learn a positive-definite curvature matrix and apply an inverse root before preconditioning. Recently, non-diagonal training methods, such as Shampoo, have gained significant attention; however, they remain computationally inefficient and are limited to specific types of curvature information due to the costly matrix root computation via matrix decomposition. To address this, we propose a Riemannian optimization approach that dynamically adapts spectral-factorized positive-definite curvature estimates, enabling the efficient application of ar","authors_text":"Felix Dangel, Juhan Bae, Richard E. Turner, Roger B. Grosse, Runa Eschenhagen, Wu Lin","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"stat.ML","submitted_at":"2025-02-10T09:07:04Z","title":"Spectral-factorized Positive-definite Curvature Learning for NN Training"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06268","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:b2f5dd9771169f9092cbfe69574442eb20e2fa7dba13be28d7be1e732cdfa1fb","target":"record","created_at":"2026-07-05T10:40:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"87047074fc315447bc29e9ef0202f65c358704ea42a608e28f0c3e4f0b335054","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"stat.ML","submitted_at":"2025-02-10T09:07:04Z","title_canon_sha256":"10e171fd4d5e3aa5788b0d82b18bac139c7e018eaed62f59d7a15c715d5c8f83"},"schema_version":"1.0","source":{"id":"2502.06268","kind":"arxiv","version":3}},"canonical_sha256":"f964548eeac5f7bf883a1f4531e4288f4052e9da7b49fde0b94135b24c84e5d2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f964548eeac5f7bf883a1f4531e4288f4052e9da7b49fde0b94135b24c84e5d2","first_computed_at":"2026-07-05T10:40:59.896903Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:40:59.896903Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5MFlidflo8sgBxlYljLSSK+2ULXYl5Q+biK/jSNY+6WXUaKEPMkXmkhUH3k7LgtJFGji47AJdNWJ2l9JbXCGBw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:40:59.897412Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.06268","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b2f5dd9771169f9092cbfe69574442eb20e2fa7dba13be28d7be1e732cdfa1fb","sha256:bcfdf40713029623753ac514666831410e54c90e86e83508294f4604ec6ab428"],"state_sha256":"bc422b3eae6fe91cc9a8548baa3a304103371263db528d209fe9f2e52e0f7778"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6TiWSoAkZu0uzV0CN97vklJ7CLCt/WaQ/QZ4NhqpxoBO+Ai1KhuTTmhHkHwE/i05RIiX7KpJHF/9H5JPVmLkAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T11:36:32.884818Z","bundle_sha256":"bc6ab7905d2ed6510a382d4e2212fea047778bfe5d8515f6be230cc47983e39a"}}