{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:KBWDWCHM5E3NMDMW3XDOBEATVQ","short_pith_number":"pith:KBWDWCHM","schema_version":"1.0","canonical_sha256":"506c3b08ece936d60d96ddc6e09013ac2eb9e2ea37303dc7f9d9671f25b2264d","source":{"kind":"arxiv","id":"2404.05728","version":6},"attestation_state":"computed","paper":{"title":"An Empirical Study of $\\mu$P Learning Rate Transfer","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lucas Lingle","submitted_at":"2024-04-08T17:59:44Z","abstract_excerpt":"Deep learning models have become a cornerstone of modern AI research, yet their initializations and learning rates may at times be set in an opaque or ad-hoc fashion due to the high cost of hyperparameter sweeps. The $\\mu$-Parameterization ($\\mu$P) offers a possible solution to this challenge, yielding scaling rules for model initialization and learning rates while reportedly enabling zero-shot hyperparameter transfer from small to large models. Despite its evident promise, the $\\mu$P method is not yet widely adopted, perhaps due to higher implementation complexity, many variations, or complex"},"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":"2404.05728","kind":"arxiv","version":6},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-04-08T17:59:44Z","cross_cats_sorted":[],"title_canon_sha256":"41cd8e9bbe51d289398d5800d369e2232d03b8412f0c9f62a7d5d3d2bbb6533d","abstract_canon_sha256":"2676b6f7686a20ff17b6b184c01cbc11e048027295ed1f68a98899fab80778a2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:14:02.834911Z","signature_b64":"FjGBw38xzO8yfqSBogVtpscjgAtgbwOsaMoamTSv77mNxn8MBwPgEXt+hLOljzY9XElGlCM0395E+Bcnzb+6Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"506c3b08ece936d60d96ddc6e09013ac2eb9e2ea37303dc7f9d9671f25b2264d","last_reissued_at":"2026-07-05T10:14:02.834418Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:14:02.834418Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"An Empirical Study of $\\mu$P Learning Rate Transfer","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lucas Lingle","submitted_at":"2024-04-08T17:59:44Z","abstract_excerpt":"Deep learning models have become a cornerstone of modern AI research, yet their initializations and learning rates may at times be set in an opaque or ad-hoc fashion due to the high cost of hyperparameter sweeps. The $\\mu$-Parameterization ($\\mu$P) offers a possible solution to this challenge, yielding scaling rules for model initialization and learning rates while reportedly enabling zero-shot hyperparameter transfer from small to large models. Despite its evident promise, the $\\mu$P method is not yet widely adopted, perhaps due to higher implementation complexity, many variations, or complex"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.05728","kind":"arxiv","version":6},"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/2404.05728/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":"2404.05728","created_at":"2026-07-05T10:14:02.834480+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.05728v6","created_at":"2026-07-05T10:14:02.834480+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.05728","created_at":"2026-07-05T10:14:02.834480+00:00"},{"alias_kind":"pith_short_12","alias_value":"KBWDWCHM5E3N","created_at":"2026-07-05T10:14:02.834480+00:00"},{"alias_kind":"pith_short_16","alias_value":"KBWDWCHM5E3NMDMW","created_at":"2026-07-05T10:14:02.834480+00:00"},{"alias_kind":"pith_short_8","alias_value":"KBWDWCHM","created_at":"2026-07-05T10:14:02.834480+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20299","citing_title":"Statistical Properties of Training & Generalization","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20299","citing_title":"Statistical Properties of Training & Generalization","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27077","citing_title":"Learning Rate Transfer in Normalized Transformers","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ","json":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ.json","graph_json":"https://pith.science/api/pith-number/KBWDWCHM5E3NMDMW3XDOBEATVQ/graph.json","events_json":"https://pith.science/api/pith-number/KBWDWCHM5E3NMDMW3XDOBEATVQ/events.json","paper":"https://pith.science/paper/KBWDWCHM"},"agent_actions":{"view_html":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ","download_json":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ.json","view_paper":"https://pith.science/paper/KBWDWCHM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.05728&json=true","fetch_graph":"https://pith.science/api/pith-number/KBWDWCHM5E3NMDMW3XDOBEATVQ/graph.json","fetch_events":"https://pith.science/api/pith-number/KBWDWCHM5E3NMDMW3XDOBEATVQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ/action/storage_attestation","attest_author":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ/action/author_attestation","sign_citation":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ/action/citation_signature","submit_replication":"https://pith.science/pith/KBWDWCHM5E3NMDMW3XDOBEATVQ/action/replication_record"}},"created_at":"2026-07-05T10:14:02.834480+00:00","updated_at":"2026-07-05T10:14:02.834480+00:00"}