{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:HIOQY2MQK7IJHZCQAEX2GI3FB6","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":"5e2d65eb604fdc11e9c8d6dd14b69356b4d6c62e1a1457b28e0abdf459596d4b","cross_cats_sorted":["math.OC","math.PR","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-04-22T15:56:43Z","title_canon_sha256":"15ba5bcc2ef2b2adc6dd1619d13c17823e219ab349a2691801361fb269e42912"},"schema_version":"1.0","source":{"id":"2204.10782","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2204.10782","created_at":"2026-07-05T04:17:01Z"},{"alias_kind":"arxiv_version","alias_value":"2204.10782v1","created_at":"2026-07-05T04:17:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2204.10782","created_at":"2026-07-05T04:17:01Z"},{"alias_kind":"pith_short_12","alias_value":"HIOQY2MQK7IJ","created_at":"2026-07-05T04:17:01Z"},{"alias_kind":"pith_short_16","alias_value":"HIOQY2MQK7IJHZCQ","created_at":"2026-07-05T04:17:01Z"},{"alias_kind":"pith_short_8","alias_value":"HIOQY2MQ","created_at":"2026-07-05T04:17:01Z"}],"graph_snapshots":[{"event_id":"sha256:01906ecc6b71e7be69c5da355a5f383b6f9516651c8ae15353f5ba5eed5c0859","target":"graph","created_at":"2026-07-05T04:17:01Z","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/2204.10782/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We study the optimization of wide neural networks (NNs) via gradient flow (GF) in setups that allow feature learning while admitting non-asymptotic global convergence guarantees. First, for wide shallow NNs under the mean-field scaling and with a general class of activation functions, we prove that when the input dimension is no less than the size of the training set, the training loss converges to zero at a linear rate under GF. Building upon this analysis, we study a model of wide multi-layer NNs whose second-to-last layer is trained via GF, for which we also prove a linear-rate convergence ","authors_text":"Eric Vanden-Eijnden, Joan Bruna, Zhengdao Chen","cross_cats":["math.OC","math.PR","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-04-22T15:56:43Z","title":"On Feature Learning in Neural Networks with Global Convergence Guarantees"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2204.10782","kind":"arxiv","version":1},"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":"sha256:71003b939f48295f98f68ab1719e12112976c84219fa6ed7c7ba656a30d4eac1","target":"record","created_at":"2026-07-05T04:17:01Z","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":"5e2d65eb604fdc11e9c8d6dd14b69356b4d6c62e1a1457b28e0abdf459596d4b","cross_cats_sorted":["math.OC","math.PR","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-04-22T15:56:43Z","title_canon_sha256":"15ba5bcc2ef2b2adc6dd1619d13c17823e219ab349a2691801361fb269e42912"},"schema_version":"1.0","source":{"id":"2204.10782","kind":"arxiv","version":1}},"canonical_sha256":"3a1d0c699057d093e450012fa323650f98763b6460d135e7619ae936a2b1b76d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3a1d0c699057d093e450012fa323650f98763b6460d135e7619ae936a2b1b76d","first_computed_at":"2026-07-05T04:17:01.051107Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:17:01.051107Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"T+5MOQNt6BFB1xSucBrHNnWss1P1RtMeHPLmIciuU81PCgW26eqpwCCUoze3GNIieLxVUL9SI5UY0sdAJx8eDg==","signature_status":"signed_v1","signed_at":"2026-07-05T04:17:01.051611Z","signed_message":"canonical_sha256_bytes"},"source_id":"2204.10782","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:71003b939f48295f98f68ab1719e12112976c84219fa6ed7c7ba656a30d4eac1","sha256:01906ecc6b71e7be69c5da355a5f383b6f9516651c8ae15353f5ba5eed5c0859"],"state_sha256":"c9ea32444cf231f64556084741670fa4b93e96291463a1295fba9ab967d0eb37"}