{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4TF45KASUQEBVXQWXCOLUDG3M7","short_pith_number":"pith:4TF45KAS","schema_version":"1.0","canonical_sha256":"e4cbcea812a4081ade16b89cba0cdb67ce76799dbf1190d9d7b41e9a8bfda987","source":{"kind":"arxiv","id":"2606.21158","version":1},"attestation_state":"computed","paper":{"title":"Dead-Direction Signatures: A Cheap Spectral Reading of Singular Complexity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"P. J. Narayanan, Tejas Pradeep Shirodkar","submitted_at":"2026-06-19T06:49:09Z","abstract_excerpt":"Singular learning theory characterises the complexity of a deep network through the geometry of its loss singularities. The local learning coefficient (LLC), the standard estimator of Watanabe's real log canonical threshold (RLCT, $\\lambda$), reads this geometry as an integrated Bayesian scalar through SGLD, which needs per-task calibration and $10^4$-$10^6$ forward-backward passes per checkpoint. We introduce Dead-Direction Signatures (DDS), a family of cheap closed-form spectral readings of singular structure: each reads a network's activation matrix or per-sample-gradient Fisher-Gram at a c"},"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":"2606.21158","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-19T06:49:09Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"3fa4a3bbac440d0140842c5912742db6298d4ac6f5c0102d66a97fdd1845b498","abstract_canon_sha256":"adaabc559b299cca31cf2b80843b1a5986aaedf4c2db561d1c72891adf4c7c07"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T01:12:31.901652Z","signature_b64":"F6helsik+/fDR/L5+RygvO12V5MBCEawG5AkK04o1pWm796QwlEg4rGkzqBYVwp3nU/LzyNlz6uyaXCNIRwXCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4cbcea812a4081ade16b89cba0cdb67ce76799dbf1190d9d7b41e9a8bfda987","last_reissued_at":"2026-06-23T01:12:31.901177Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T01:12:31.901177Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dead-Direction Signatures: A Cheap Spectral Reading of Singular Complexity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"P. J. Narayanan, Tejas Pradeep Shirodkar","submitted_at":"2026-06-19T06:49:09Z","abstract_excerpt":"Singular learning theory characterises the complexity of a deep network through the geometry of its loss singularities. The local learning coefficient (LLC), the standard estimator of Watanabe's real log canonical threshold (RLCT, $\\lambda$), reads this geometry as an integrated Bayesian scalar through SGLD, which needs per-task calibration and $10^4$-$10^6$ forward-backward passes per checkpoint. We introduce Dead-Direction Signatures (DDS), a family of cheap closed-form spectral readings of singular structure: each reads a network's activation matrix or per-sample-gradient Fisher-Gram at a c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.21158","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/2606.21158/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":"2606.21158","created_at":"2026-06-23T01:12:31.901251+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.21158v1","created_at":"2026-06-23T01:12:31.901251+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.21158","created_at":"2026-06-23T01:12:31.901251+00:00"},{"alias_kind":"pith_short_12","alias_value":"4TF45KASUQEB","created_at":"2026-06-23T01:12:31.901251+00:00"},{"alias_kind":"pith_short_16","alias_value":"4TF45KASUQEBVXQW","created_at":"2026-06-23T01:12:31.901251+00:00"},{"alias_kind":"pith_short_8","alias_value":"4TF45KAS","created_at":"2026-06-23T01:12:31.901251+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.29176","citing_title":"Dead-Direction Conditioners: Gauge-Equivariant Preconditioning for Deep Networks","ref_index":29,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7","json":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7.json","graph_json":"https://pith.science/api/pith-number/4TF45KASUQEBVXQWXCOLUDG3M7/graph.json","events_json":"https://pith.science/api/pith-number/4TF45KASUQEBVXQWXCOLUDG3M7/events.json","paper":"https://pith.science/paper/4TF45KAS"},"agent_actions":{"view_html":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7","download_json":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7.json","view_paper":"https://pith.science/paper/4TF45KAS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.21158&json=true","fetch_graph":"https://pith.science/api/pith-number/4TF45KASUQEBVXQWXCOLUDG3M7/graph.json","fetch_events":"https://pith.science/api/pith-number/4TF45KASUQEBVXQWXCOLUDG3M7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7/action/storage_attestation","attest_author":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7/action/author_attestation","sign_citation":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7/action/citation_signature","submit_replication":"https://pith.science/pith/4TF45KASUQEBVXQWXCOLUDG3M7/action/replication_record"}},"created_at":"2026-06-23T01:12:31.901251+00:00","updated_at":"2026-06-23T01:12:31.901251+00:00"}