{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:NCJLQUIFQZQZTGV4HT2F6X3GHH","short_pith_number":"pith:NCJLQUIF","schema_version":"1.0","canonical_sha256":"6892b851058661999abc3cf45f5f6639ee9f37ed04a66d971d6fa8e49bb540cc","source":{"kind":"arxiv","id":"1901.10159","version":1},"attestation_state":"computed","paper":{"title":"An Investigation into Neural Net Optimization via Hessian Eigenvalue Density","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Behrooz Ghorbani, Shankar Krishnan, Ying Xiao","submitted_at":"2019-01-29T08:24:10Z","abstract_excerpt":"To understand the dynamics of optimization in deep neural networks, we develop a tool to study the evolution of the entire Hessian spectrum throughout the optimization process. Using this, we study a number of hypotheses concerning smoothness, curvature, and sharpness in the deep learning literature. We then thoroughly analyze a crucial structural feature of the spectra: in non-batch normalized networks, we observe the rapid appearance of large isolated eigenvalues in the spectrum, along with a surprising concentration of the gradient in the corresponding eigenspaces. In batch normalized netwo"},"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":"1901.10159","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-29T08:24:10Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"0739269d27617185c4d4b143acfe66603c03c1ffeb0b5b30fb86539ec620dc17","abstract_canon_sha256":"9c40bb2b5c82efcc953f4f50419be2ecf645080f933eb2a91febb7839468f54d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:55:15.996530Z","signature_b64":"+q2HNMPgi3ohBSeXgDp8kON1zo1ymktymhydPxQp7W9pk/OPPhn/zVD6pwPUl7Fik87R5KpJtfKJPNAfkIk0Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6892b851058661999abc3cf45f5f6639ee9f37ed04a66d971d6fa8e49bb540cc","last_reissued_at":"2026-05-17T23:55:15.995900Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:55:15.995900Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"An Investigation into Neural Net Optimization via Hessian Eigenvalue Density","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Behrooz Ghorbani, Shankar Krishnan, Ying Xiao","submitted_at":"2019-01-29T08:24:10Z","abstract_excerpt":"To understand the dynamics of optimization in deep neural networks, we develop a tool to study the evolution of the entire Hessian spectrum throughout the optimization process. Using this, we study a number of hypotheses concerning smoothness, curvature, and sharpness in the deep learning literature. We then thoroughly analyze a crucial structural feature of the spectra: in non-batch normalized networks, we observe the rapid appearance of large isolated eigenvalues in the spectrum, along with a surprising concentration of the gradient in the corresponding eigenspaces. In batch normalized netwo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1901.10159","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":""},"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":"1901.10159","created_at":"2026-05-17T23:55:15.996012+00:00"},{"alias_kind":"arxiv_version","alias_value":"1901.10159v1","created_at":"2026-05-17T23:55:15.996012+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1901.10159","created_at":"2026-05-17T23:55:15.996012+00:00"},{"alias_kind":"pith_short_12","alias_value":"NCJLQUIFQZQZ","created_at":"2026-05-18T12:33:24.271573+00:00"},{"alias_kind":"pith_short_16","alias_value":"NCJLQUIFQZQZTGV4","created_at":"2026-05-18T12:33:24.271573+00:00"},{"alias_kind":"pith_short_8","alias_value":"NCJLQUIF","created_at":"2026-05-18T12:33:24.271573+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2605.23476","citing_title":"Non-normal spectral signatures of instability in neural network training dynamics","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"1907.10732","citing_title":"Hessian based analysis of SGD for Deep Nets: Dynamics and Generalization","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2102.01293","citing_title":"Scaling Laws for Transfer","ref_index":76,"is_internal_anchor":true},{"citing_arxiv_id":"2112.00861","citing_title":"A General Language Assistant as a Laboratory for Alignment","ref_index":106,"is_internal_anchor":false},{"citing_arxiv_id":"2207.05221","citing_title":"Language Models (Mostly) Know What They Know","ref_index":164,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH","json":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH.json","graph_json":"https://pith.science/api/pith-number/NCJLQUIFQZQZTGV4HT2F6X3GHH/graph.json","events_json":"https://pith.science/api/pith-number/NCJLQUIFQZQZTGV4HT2F6X3GHH/events.json","paper":"https://pith.science/paper/NCJLQUIF"},"agent_actions":{"view_html":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH","download_json":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH.json","view_paper":"https://pith.science/paper/NCJLQUIF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1901.10159&json=true","fetch_graph":"https://pith.science/api/pith-number/NCJLQUIFQZQZTGV4HT2F6X3GHH/graph.json","fetch_events":"https://pith.science/api/pith-number/NCJLQUIFQZQZTGV4HT2F6X3GHH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH/action/storage_attestation","attest_author":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH/action/author_attestation","sign_citation":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH/action/citation_signature","submit_replication":"https://pith.science/pith/NCJLQUIFQZQZTGV4HT2F6X3GHH/action/replication_record"}},"created_at":"2026-05-17T23:55:15.996012+00:00","updated_at":"2026-05-17T23:55:15.996012+00:00"}