{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YJBKOT33SAPL22WE4ZIVW2MUWA","short_pith_number":"pith:YJBKOT33","schema_version":"1.0","canonical_sha256":"c242a74f7b901ebd6ac4e6515b6994b02dac4eb005ff915992d055a49b92db22","source":{"kind":"arxiv","id":"2504.18208","version":2},"attestation_state":"computed","paper":{"title":"Ultra-fast feature learning for the training of two-layer neural networks in the two-timescale regime","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"\\'ENS-PSL), Fran\\c{c}ois-Xavier Vialard (LIGM), Gabriel Peyr\\'e (CNRS, Rapha\\\"el Barboni (\\'ENS-PSL)","submitted_at":"2025-04-25T09:40:10Z","abstract_excerpt":"We study the convergence of gradient methods for the training of mean-field single-hidden-layer neural networks with square loss. For this high-dimensional and non-convex optimization problem, most known convergence results are either qualitative or rely on a neural tangent kernel analysis where nonlinear representations of the data are fixed. Using that this problem belongs to the class of separable nonlinear least squares problems, we consider here a Variable Projection (VarPro) or two-timescale learning algorithm, thereby eliminating the linear variables and reducing the learning problem to"},"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":"2504.18208","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-04-25T09:40:10Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"8449ea9f2249b6eefcee4bd2b37c67cf7a685cf89834933c9925d6f1657431eb","abstract_canon_sha256":"ef1dc3a52faa45d0cfae2d948ea9ad317b90e63369a607e62d6d0f83f25d6b47"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:40:34.066384Z","signature_b64":"J2AxVaro0G3NMp2olApklTRx6rSnSLW7YK7bcjThIzjxGcsOjzRcj74/9RIYhhfIpfKYaVcttTgE3crR5gNtDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c242a74f7b901ebd6ac4e6515b6994b02dac4eb005ff915992d055a49b92db22","last_reissued_at":"2026-07-05T11:40:34.065961Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:40:34.065961Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Ultra-fast feature learning for the training of two-layer neural networks in the two-timescale regime","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"\\'ENS-PSL), Fran\\c{c}ois-Xavier Vialard (LIGM), Gabriel Peyr\\'e (CNRS, Rapha\\\"el Barboni (\\'ENS-PSL)","submitted_at":"2025-04-25T09:40:10Z","abstract_excerpt":"We study the convergence of gradient methods for the training of mean-field single-hidden-layer neural networks with square loss. For this high-dimensional and non-convex optimization problem, most known convergence results are either qualitative or rely on a neural tangent kernel analysis where nonlinear representations of the data are fixed. Using that this problem belongs to the class of separable nonlinear least squares problems, we consider here a Variable Projection (VarPro) or two-timescale learning algorithm, thereby eliminating the linear variables and reducing the learning problem to"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.18208","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/2504.18208/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":"2504.18208","created_at":"2026-07-05T11:40:34.066016+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.18208v2","created_at":"2026-07-05T11:40:34.066016+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.18208","created_at":"2026-07-05T11:40:34.066016+00:00"},{"alias_kind":"pith_short_12","alias_value":"YJBKOT33SAPL","created_at":"2026-07-05T11:40:34.066016+00:00"},{"alias_kind":"pith_short_16","alias_value":"YJBKOT33SAPL22WE","created_at":"2026-07-05T11:40:34.066016+00:00"},{"alias_kind":"pith_short_8","alias_value":"YJBKOT33","created_at":"2026-07-05T11:40:34.066016+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.15530","citing_title":"Rethinking Neural Network Learning Rates: A Stackelberg Perspective","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15530","citing_title":"Rethinking Neural Network Learning Rates: A Stackelberg Perspective","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2510.04606","citing_title":"Closed-Form Last Layer Optimization","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA","json":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA.json","graph_json":"https://pith.science/api/pith-number/YJBKOT33SAPL22WE4ZIVW2MUWA/graph.json","events_json":"https://pith.science/api/pith-number/YJBKOT33SAPL22WE4ZIVW2MUWA/events.json","paper":"https://pith.science/paper/YJBKOT33"},"agent_actions":{"view_html":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA","download_json":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA.json","view_paper":"https://pith.science/paper/YJBKOT33","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.18208&json=true","fetch_graph":"https://pith.science/api/pith-number/YJBKOT33SAPL22WE4ZIVW2MUWA/graph.json","fetch_events":"https://pith.science/api/pith-number/YJBKOT33SAPL22WE4ZIVW2MUWA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA/action/storage_attestation","attest_author":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA/action/author_attestation","sign_citation":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA/action/citation_signature","submit_replication":"https://pith.science/pith/YJBKOT33SAPL22WE4ZIVW2MUWA/action/replication_record"}},"created_at":"2026-07-05T11:40:34.066016+00:00","updated_at":"2026-07-05T11:40:34.066016+00:00"}