{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:MJ6PJPKRHTF7VSRBYEMF5HQYKC","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":"c0436c978d7d841b43e8c67a2b355e1106b32e3c618722cf48f51e36b41cb646","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-07-24T17:58:42Z","title_canon_sha256":"cb30de77e7f132945e4bb9bb2a8bc9b0e2950fa1b1d48d845fce746845bd2c52"},"schema_version":"1.0","source":{"id":"2407.17465","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.17465","created_at":"2026-07-05T09:59:17Z"},{"alias_kind":"arxiv_version","alias_value":"2407.17465v3","created_at":"2026-07-05T09:59:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.17465","created_at":"2026-07-05T09:59:17Z"},{"alias_kind":"pith_short_12","alias_value":"MJ6PJPKRHTF7","created_at":"2026-07-05T09:59:17Z"},{"alias_kind":"pith_short_16","alias_value":"MJ6PJPKRHTF7VSRB","created_at":"2026-07-05T09:59:17Z"},{"alias_kind":"pith_short_8","alias_value":"MJ6PJPKR","created_at":"2026-07-05T09:59:17Z"}],"graph_snapshots":[{"event_id":"sha256:eb572757ed118b2acab72677aa4859199281c28e7dbb512079a1c9eb1f75fabc","target":"graph","created_at":"2026-07-05T09:59:17Z","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/2407.17465/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The Maximal Update Parametrization ($\\mu$P) aims to make the optimal hyperparameters (HPs) of a model independent of its size, allowing them to be swept using a cheap proxy model rather than the full-size target model. We present a new scheme, u-$\\mu$P, which improves upon $\\mu$P by combining it with Unit Scaling, a method for designing models that makes them easy to train in low-precision. The two techniques have a natural affinity: $\\mu$P ensures that the scale of activations is independent of model size, and Unit Scaling ensures that activations, weights and gradients begin training with a ","authors_text":"Andres Felipe Cruz-Salinas, Bj\\\"orn Deiseroth, Carlo Luschi, Charlie Blake, Constantin Eichenberg, Douglas Orr, Josef Dean, Lukas Balles, Luke Y. Prince, Samuel Weinbach","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-07-24T17:58:42Z","title":"u-$\\mu$P: The Unit-Scaled Maximal Update Parametrization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.17465","kind":"arxiv","version":3},"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":"sha256:28d022bf68e81d3e0f3ce570280bcda70f1974a0fbf34760e16eb818e268638d","target":"record","created_at":"2026-07-05T09:59:17Z","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":"c0436c978d7d841b43e8c67a2b355e1106b32e3c618722cf48f51e36b41cb646","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-07-24T17:58:42Z","title_canon_sha256":"cb30de77e7f132945e4bb9bb2a8bc9b0e2950fa1b1d48d845fce746845bd2c52"},"schema_version":"1.0","source":{"id":"2407.17465","kind":"arxiv","version":3}},"canonical_sha256":"627cf4bd513ccbfaca21c1185e9e18509141efa7ad1896b28785b3a621752851","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"627cf4bd513ccbfaca21c1185e9e18509141efa7ad1896b28785b3a621752851","first_computed_at":"2026-07-05T09:59:17.796858Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:59:17.796858Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"QbPQ2Ya6B/iJ5D3QIulDnjUde1XF3VidQ0sIYuCDfpX4djIGOEqPgWC5jhsPQM48DF5O7Jnld+KXS2WVOPYpBA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:59:17.797340Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.17465","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:28d022bf68e81d3e0f3ce570280bcda70f1974a0fbf34760e16eb818e268638d","sha256:eb572757ed118b2acab72677aa4859199281c28e7dbb512079a1c9eb1f75fabc"],"state_sha256":"7a6924d21a19ad8255f02ddbf1b8efaeed0fd5730a138f35dccfa088eef8499a"}