{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:VYTKFDUTU36YRB5UQQLEV64VQN","short_pith_number":"pith:VYTKFDUT","canonical_record":{"source":{"id":"2605.08352","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-08T18:02:37Z","cross_cats_sorted":["math.PR","stat.ML"],"title_canon_sha256":"84e6314d565c0160b7e80e372d99f57b5d903d62c81c316e3a99dbe81d497087","abstract_canon_sha256":"99515b2b2f5d1f52f3697959431d65f6a38e26548e2d9251e7f985ffcf779346"},"schema_version":"1.0"},"canonical_sha256":"ae26a28e93a6fd8887b484164afb95836f906ed86c316ed3060a032c06ea99d2","source":{"kind":"arxiv","id":"2605.08352","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.08352","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"arxiv_version","alias_value":"2605.08352v2","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.08352","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"pith_short_12","alias_value":"VYTKFDUTU36Y","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"pith_short_16","alias_value":"VYTKFDUTU36YRB5U","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"pith_short_8","alias_value":"VYTKFDUT","created_at":"2026-05-21T02:05:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:VYTKFDUTU36YRB5UQQLEV64VQN","target":"record","payload":{"canonical_record":{"source":{"id":"2605.08352","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-08T18:02:37Z","cross_cats_sorted":["math.PR","stat.ML"],"title_canon_sha256":"84e6314d565c0160b7e80e372d99f57b5d903d62c81c316e3a99dbe81d497087","abstract_canon_sha256":"99515b2b2f5d1f52f3697959431d65f6a38e26548e2d9251e7f985ffcf779346"},"schema_version":"1.0"},"canonical_sha256":"ae26a28e93a6fd8887b484164afb95836f906ed86c316ed3060a032c06ea99d2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T02:05:04.806763Z","signature_b64":"oNuMAwShhWdDD/LTn28V3oBWnNK+uQySuRB/IBrqHT6DHlzG9VNNBnkt7SDJXkjatXUhSEbDOwd8N/juG321Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ae26a28e93a6fd8887b484164afb95836f906ed86c316ed3060a032c06ea99d2","last_reissued_at":"2026-05-21T02:05:04.806156Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T02:05:04.806156Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.08352","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-21T02:05:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2fKG/Xwy87rePbmWSC142ixKhTcS9tbbgDkiAFURzmB4+laNkLhgwIAqyZ7y5zkWBBiqArGjSlIB9rEfDwR7AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T21:18:20.609135Z"},"content_sha256":"b2d6f3faf1234f0d9d98beddea4e0286932d5c967ca48a15c8d43863f550bfcf","schema_version":"1.0","event_id":"sha256:b2d6f3faf1234f0d9d98beddea4e0286932d5c967ca48a15c8d43863f550bfcf"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:VYTKFDUTU36YRB5UQQLEV64VQN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Convergence Analysis of Newton's Method for Neural Networks in the Overparameterized Limit","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Regularized Newton's method for neural networks converges exponentially to zero loss in the infinite-width limit uniformly across frequencies.","cross_cats":["math.PR","stat.ML"],"primary_cat":"cs.LG","authors_text":"Justin Sirignano, Konstantinos Spiliopoulos, Konstantin Riedl","submitted_at":"2026-05-08T18:02:37Z","abstract_excerpt":"A convergence analysis is developed for the regularized Newton method for training neural networks (NNs) in the overparameterized limit. As the number of hidden units tends to infinity, the NN training dynamics converge in probability to the solution of a deterministic limit equation involving a ``Newton neural tangent kernel'' (NNTK). Explicit rates characterizing this convergence are provided and, in the infinite-width limit, we prove that the NN converges exponentially fast to the target data (i.e., a global minimizer with zero loss). We show that this convergence is uniform across the freq"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"in the infinite-width limit, we prove that the NN converges exponentially fast to the target data (i.e., a global minimizer with zero loss). We show that this convergence is uniform across the frequency spectrum, addressing the spectral bias inherent in gradient descent.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The regularization parameter can be chosen via a scaling formula that vanishes at a suitable rate as the number of hidden units grows, ensuring the regularized Hessian remains positive definite for sufficiently large widths during training.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"In the infinite-width limit, regularized Newton's method for neural networks converges exponentially to global minimizers with uniform rates across the frequency spectrum using the Newton neural tangent kernel.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Regularized Newton's method for neural networks converges exponentially to zero loss in the infinite-width limit uniformly across frequencies.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"6c528ae4f793fba40d4a3333a4e9bbaf5132991fc0982f5f6d765c9871559fee"},"source":{"id":"2605.08352","kind":"arxiv","version":2},"verdict":{"id":"ba36840e-a056-44b7-9242-ae235e27738d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-12T01:11:15.265890Z","strongest_claim":"in the infinite-width limit, we prove that the NN converges exponentially fast to the target data (i.e., a global minimizer with zero loss). We show that this convergence is uniform across the frequency spectrum, addressing the spectral bias inherent in gradient descent.","one_line_summary":"In the infinite-width limit, regularized Newton's method for neural networks converges exponentially to global minimizers with uniform rates across the frequency spectrum using the Newton neural tangent kernel.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The regularization parameter can be chosen via a scaling formula that vanishes at a suitable rate as the number of hidden units grows, ensuring the regularized Hessian remains positive definite for sufficiently large widths during training.","pith_extraction_headline":"Regularized Newton's method for neural networks converges exponentially to zero loss in the infinite-width limit uniformly across frequencies."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.08352/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T09:42:04.297188Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-20T04:38:29.340051Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T15:01:18.289185Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T11:12:38.463148Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"7ddf18a1935bb8f7d48151bde0a7f7693bf5193d09fcfd05a8c3a45a77105d0d"},"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":"ba36840e-a056-44b7-9242-ae235e27738d"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-21T02:05:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gMgWwu1TfDjxMIf3ha8RBENRs9ECXkB32hwbBrx7FyWYljneHnVvkKgM58vhPM89mCdI5GfnV7RSXHMLbpwGDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T21:18:20.610100Z"},"content_sha256":"db2d1104f37e42d8f63d9be36301b0e33c1568de3529de3edc76b59dc5ba340c","schema_version":"1.0","event_id":"sha256:db2d1104f37e42d8f63d9be36301b0e33c1568de3529de3edc76b59dc5ba340c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VYTKFDUTU36YRB5UQQLEV64VQN/bundle.json","state_url":"https://pith.science/pith/VYTKFDUTU36YRB5UQQLEV64VQN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VYTKFDUTU36YRB5UQQLEV64VQN/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":"2026-05-25T21:18:20Z","links":{"resolver":"https://pith.science/pith/VYTKFDUTU36YRB5UQQLEV64VQN","bundle":"https://pith.science/pith/VYTKFDUTU36YRB5UQQLEV64VQN/bundle.json","state":"https://pith.science/pith/VYTKFDUTU36YRB5UQQLEV64VQN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VYTKFDUTU36YRB5UQQLEV64VQN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:VYTKFDUTU36YRB5UQQLEV64VQN","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":"99515b2b2f5d1f52f3697959431d65f6a38e26548e2d9251e7f985ffcf779346","cross_cats_sorted":["math.PR","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-08T18:02:37Z","title_canon_sha256":"84e6314d565c0160b7e80e372d99f57b5d903d62c81c316e3a99dbe81d497087"},"schema_version":"1.0","source":{"id":"2605.08352","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.08352","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"arxiv_version","alias_value":"2605.08352v2","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.08352","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"pith_short_12","alias_value":"VYTKFDUTU36Y","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"pith_short_16","alias_value":"VYTKFDUTU36YRB5U","created_at":"2026-05-21T02:05:04Z"},{"alias_kind":"pith_short_8","alias_value":"VYTKFDUT","created_at":"2026-05-21T02:05:04Z"}],"graph_snapshots":[{"event_id":"sha256:db2d1104f37e42d8f63d9be36301b0e33c1568de3529de3edc76b59dc5ba340c","target":"graph","created_at":"2026-05-21T02:05:04Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"in the infinite-width limit, we prove that the NN converges exponentially fast to the target data (i.e., a global minimizer with zero loss). We show that this convergence is uniform across the frequency spectrum, addressing the spectral bias inherent in gradient descent."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The regularization parameter can be chosen via a scaling formula that vanishes at a suitable rate as the number of hidden units grows, ensuring the regularized Hessian remains positive definite for sufficiently large widths during training."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"In the infinite-width limit, regularized Newton's method for neural networks converges exponentially to global minimizers with uniform rates across the frequency spectrum using the Newton neural tangent kernel."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Regularized Newton's method for neural networks converges exponentially to zero loss in the infinite-width limit uniformly across frequencies."}],"snapshot_sha256":"6c528ae4f793fba40d4a3333a4e9bbaf5132991fc0982f5f6d765c9871559fee"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-20T09:42:04.297188Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-20T04:38:29.340051Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T15:01:18.289185Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T11:12:38.463148Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.08352/integrity.json","findings":[],"snapshot_sha256":"7ddf18a1935bb8f7d48151bde0a7f7693bf5193d09fcfd05a8c3a45a77105d0d","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"A convergence analysis is developed for the regularized Newton method for training neural networks (NNs) in the overparameterized limit. As the number of hidden units tends to infinity, the NN training dynamics converge in probability to the solution of a deterministic limit equation involving a ``Newton neural tangent kernel'' (NNTK). Explicit rates characterizing this convergence are provided and, in the infinite-width limit, we prove that the NN converges exponentially fast to the target data (i.e., a global minimizer with zero loss). We show that this convergence is uniform across the freq","authors_text":"Justin Sirignano, Konstantinos Spiliopoulos, Konstantin Riedl","cross_cats":["math.PR","stat.ML"],"headline":"Regularized Newton's method for neural networks converges exponentially to zero loss in the infinite-width limit uniformly across frequencies.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-08T18:02:37Z","title":"Convergence Analysis of Newton's Method for Neural Networks in the Overparameterized Limit"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.08352","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-12T01:11:15.265890Z","id":"ba36840e-a056-44b7-9242-ae235e27738d","model_set":{"reader":"grok-4.3"},"one_line_summary":"In the infinite-width limit, regularized Newton's method for neural networks converges exponentially to global minimizers with uniform rates across the frequency spectrum using the Newton neural tangent kernel.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Regularized Newton's method for neural networks converges exponentially to zero loss in the infinite-width limit uniformly across frequencies.","strongest_claim":"in the infinite-width limit, we prove that the NN converges exponentially fast to the target data (i.e., a global minimizer with zero loss). We show that this convergence is uniform across the frequency spectrum, addressing the spectral bias inherent in gradient descent.","weakest_assumption":"The regularization parameter can be chosen via a scaling formula that vanishes at a suitable rate as the number of hidden units grows, ensuring the regularized Hessian remains positive definite for sufficiently large widths during training."}},"verdict_id":"ba36840e-a056-44b7-9242-ae235e27738d"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:b2d6f3faf1234f0d9d98beddea4e0286932d5c967ca48a15c8d43863f550bfcf","target":"record","created_at":"2026-05-21T02:05:04Z","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":"99515b2b2f5d1f52f3697959431d65f6a38e26548e2d9251e7f985ffcf779346","cross_cats_sorted":["math.PR","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-08T18:02:37Z","title_canon_sha256":"84e6314d565c0160b7e80e372d99f57b5d903d62c81c316e3a99dbe81d497087"},"schema_version":"1.0","source":{"id":"2605.08352","kind":"arxiv","version":2}},"canonical_sha256":"ae26a28e93a6fd8887b484164afb95836f906ed86c316ed3060a032c06ea99d2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ae26a28e93a6fd8887b484164afb95836f906ed86c316ed3060a032c06ea99d2","first_computed_at":"2026-05-21T02:05:04.806156Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-21T02:05:04.806156Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oNuMAwShhWdDD/LTn28V3oBWnNK+uQySuRB/IBrqHT6DHlzG9VNNBnkt7SDJXkjatXUhSEbDOwd8N/juG321Dg==","signature_status":"signed_v1","signed_at":"2026-05-21T02:05:04.806763Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.08352","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b2d6f3faf1234f0d9d98beddea4e0286932d5c967ca48a15c8d43863f550bfcf","sha256:db2d1104f37e42d8f63d9be36301b0e33c1568de3529de3edc76b59dc5ba340c"],"state_sha256":"68749cacb122bfac32cd9abaaf3f4393c2ebeb0142b20d7effe7e01f8ec5f6c2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"c2v84SvmLJvyPPSsbti8BxANQHB/hC4roQBxLg3MBWqBv7w52pdXKW2zpJIHa18IzEgle0SfCzVExwtBP1lcBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T21:18:20.613582Z","bundle_sha256":"78a308805f25ff1a8418dcb8315de3e8e0dc93d901c60325e7194de933ce82b4"}}