{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:JZDA4FQPE5DKPQHAFGUXIQKMH2","short_pith_number":"pith:JZDA4FQP","schema_version":"1.0","canonical_sha256":"4e460e160f2746a7c0e029a974414c3eb07ffe81e8234238f840af6692847278","source":{"kind":"arxiv","id":"2006.07721","version":2},"attestation_state":"computed","paper":{"title":"Beyond Random Matrix Theory for Deep Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Diego Granziol","submitted_at":"2020-06-13T21:00:30Z","abstract_excerpt":"We investigate whether the Wigner semi-circle and Marcenko-Pastur distributions, often used for deep neural network theoretical analysis, match empirically observed spectral densities. We find that even allowing for outliers, the observed spectral shapes strongly deviate from such theoretical predictions. This raises major questions about the usefulness of these models in deep learning. We further show that theoretical results, such as the layered nature of critical points, are strongly dependent on the use of the exact form of these limiting spectral densities. We consider two new classes of "},"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":"2006.07721","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-06-13T21:00:30Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"cdd9fe97a6b0916efab9a269fa0a7ed201cbc42bd7758e35f2bca50392b0a9d1","abstract_canon_sha256":"102e75a6b3b046514ce1e868f117e59ad6797ff3e8d82fdea7a90f215b4a9e8b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:28:34.738622Z","signature_b64":"4qEV7Q3jBm+LQnxdKi9O0KTXQzDtMEVcna/ojL/sa47hchmSMMkNh+0huSO+oqLmQVG4ATuG8Nf7H1oWqO11BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4e460e160f2746a7c0e029a974414c3eb07ffe81e8234238f840af6692847278","last_reissued_at":"2026-07-05T03:28:34.738151Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:28:34.738151Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Random Matrix Theory for Deep Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Diego Granziol","submitted_at":"2020-06-13T21:00:30Z","abstract_excerpt":"We investigate whether the Wigner semi-circle and Marcenko-Pastur distributions, often used for deep neural network theoretical analysis, match empirically observed spectral densities. We find that even allowing for outliers, the observed spectral shapes strongly deviate from such theoretical predictions. This raises major questions about the usefulness of these models in deep learning. We further show that theoretical results, such as the layered nature of critical points, are strongly dependent on the use of the exact form of these limiting spectral densities. We consider two new classes of "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2006.07721","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/2006.07721/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":"2006.07721","created_at":"2026-07-05T03:28:34.738207+00:00"},{"alias_kind":"arxiv_version","alias_value":"2006.07721v2","created_at":"2026-07-05T03:28:34.738207+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.07721","created_at":"2026-07-05T03:28:34.738207+00:00"},{"alias_kind":"pith_short_12","alias_value":"JZDA4FQPE5DK","created_at":"2026-07-05T03:28:34.738207+00:00"},{"alias_kind":"pith_short_16","alias_value":"JZDA4FQPE5DKPQHA","created_at":"2026-07-05T03:28:34.738207+00:00"},{"alias_kind":"pith_short_8","alias_value":"JZDA4FQP","created_at":"2026-07-05T03:28:34.738207+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.30226","citing_title":"Characterizing Optimizer-Dependent Training Dynamics Through Hessian Eigenvector Displacement and Localization","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2","json":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2.json","graph_json":"https://pith.science/api/pith-number/JZDA4FQPE5DKPQHAFGUXIQKMH2/graph.json","events_json":"https://pith.science/api/pith-number/JZDA4FQPE5DKPQHAFGUXIQKMH2/events.json","paper":"https://pith.science/paper/JZDA4FQP"},"agent_actions":{"view_html":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2","download_json":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2.json","view_paper":"https://pith.science/paper/JZDA4FQP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2006.07721&json=true","fetch_graph":"https://pith.science/api/pith-number/JZDA4FQPE5DKPQHAFGUXIQKMH2/graph.json","fetch_events":"https://pith.science/api/pith-number/JZDA4FQPE5DKPQHAFGUXIQKMH2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2/action/storage_attestation","attest_author":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2/action/author_attestation","sign_citation":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2/action/citation_signature","submit_replication":"https://pith.science/pith/JZDA4FQPE5DKPQHAFGUXIQKMH2/action/replication_record"}},"created_at":"2026-07-05T03:28:34.738207+00:00","updated_at":"2026-07-05T03:28:34.738207+00:00"}