Home › Code › zeropower_via_newtonschulz5

zeropower_via_newtonschulz5

Syntologyentry name in harvested coderead from the graph 2026-09-24

zeropower_via_newtonschulz5 appears in the code Syntology harvested for 24 papers, as 27 distinct code bodies found in 34 places (a place is one code body under one paper). At least one of them ran in 13 of the papers; 9 of the code bodies carry a behaviour fingerprint.

What this page is not. Routines are grouped here by the exact string of their function or class name. Nothing asserts that two samples named zeropower_via_newtonschulz5 do the same thing, share code, or are comparable; the name is a string, not an identity. Behaviour outputs (what a fingerprinted sample returned on the shared battery) are not in this export and are not shown here; the graph at syntology.ai holds them. "Ran" means executed on a synthesized fixture, not that the code is correct or reproduces a paper.

Samples Syntology

Syntology ran 9 of the 27 distinct code bodies named zeropower_via_newtonschulz5; 18 are unverified. One tile per status, in the site's fixed vocabulary, each code body counted once:

0ran · honoured contract
0ran · violated contract
3ran · our draft was wrong
0ran · fixture could not drive it
6ran
18unverified
9fingerprinted

Licence is a property of each copy, so it is counted per place: 8 of the 34 places are pointer only (Syntology does not serve that copy's text). This site shows no code text for any sample; every row below links to the file in its repository where the record names one.

“Ran” means the sample executed on a synthesized input; it does not mean the output is correct. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code, and those samples did run. The ran count above is every status except unverified, the same rule as each paper page.

Papers

24 papers shown of 24, newest first; 34 places in the table. A paper with no recorded date is placed by the month its arXiv id encodes, shown in the Date column as YYYY-MM (from id). One row per place: a paper whose repository defines the name more than once appears more than once, and the same code body held for several papers appears once under each, with the same status. Titles and dates are the archive's archive 2025-07-28 for papers in the archive, and the graph's for 20 papers added by Syntology. Status and fingerprint are Syntology's record of each code body; licence is recorded for each place. The File cell ends with the code body's code_sha256, Syntology's identity for that exact code: an agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

PaperDateFileStatus SyntologyLicence
Depth and Scale in the Sub-150M Regime: JugnuLM-53M vs JugnuLM-110M added by Syntology 2026-09 (from id) AltSlate-Labs/jugnu/muon.py c8c6c22862b64a4d unverified Apache-2.0 (permissive)
Squeezing More from Limited Data with Recursive Transformers added by Syntology 2026-08 (from id) serdardoesml/recursive-lm/recursive_lm/optimizer.py b7f41e9dfacd5ee3 ran fingerprinted no licence file found · pointer only
A Physical Response-and-Memory Model for Muon Optimization added by Syntology 2026-08 (from id) orange4664/bimaxwell-track3-reproduction/record_2635_submission/train_gpt_bimaxwell_2635.py c9dd31c1e98b4ebf ran · our draft was wrong fingerprinted no licence file found · pointer only
Autoregressive Boltzmann Generators added by Syntology 2026-06 (from id) danyalrehman/autobg/src/optimizers/muon.py 227822d98406033b ran fingerprinted licence not identified · pointer only
Zeta: Dual Whitening for Matrix Optimization via Coordinate-Adaptive Preconditioning added by Syntology 2026-06 (from id) AIGCodeOS/aigcode_zeta_optimizer/megatron/core/optimizer/adamuon.py 0c0438262f800460 ran fingerprinted licence not identified · pointer only
Zeta: Dual Whitening for Matrix Optimization via Coordinate-Adaptive Preconditioning added by Syntology 2026-06 (from id) AIGCodeOS/aigcode_zeta_optimizer/megatron/core/optimizer/mixmuon.py a097b474f96888d8 ran fingerprinted licence not identified · pointer only
PithTrain: A Compact and Agent-Native MoE Training System added by Syntology 2026-05 (from id) mlc-ai/pith-train/pithtrain/modules/optimizer.py f1822f08ab8bef43 ran fingerprinted Apache-2.0 (permissive)
One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs added by Syntology 2026-05 (from id) hed-ucas/Layer-wise-Learning-Rate/galore_utils/muon.py 88b19bfaad9384f8 ran · our draft was wrong fingerprinted no licence file found · pointer only
LionMuon: Alternating Spectral and Sign Descent for Efficient Training added by Syntology 2026-05 (from id) brain-lab-research/lion-muon/src/optim/muon.py 04049769e4eed44a ran fingerprinted MIT (permissive)
DynMuon: A Dynamic Spectral Shaping View of Muon added by Syntology 2026-05 (from id) fzwark/DynMuon/dynmuon/muon_reference.py 2c92c3e5770228d4 unverified MIT (permissive)
DynMuon: A Dynamic Spectral Shaping View of Muon added by Syntology 2026-05 (from id) fzwark/DynMuon/dynmuon/newton_schulz_triton.py fbcadf5bf33cd8bd unverified MIT (permissive)
A Hierarchical Language Model with Predictable Scaling Laws and Provable Benefits of Reasoning added by Syntology 2026-05 (from id) joonhyungshin/nanocontext/nanocontext/optim/muon.py 053ca60c81f1dfff unverified MIT (permissive)
Optimizer-Induced Mode Connectivity: From AdamW to Muon added by Syntology 2026-05 (from id) KellerJordan/Muon/muon.py 88b19bfaad9384f8 ran · our draft was wrong fingerprinted MIT (permissive)
Muon with Nesterov Momentum: Heavy-Tailed Noise and (Randomized) Inexact Polar Decomposition added by Syntology 2026-05 (from id) Xiaoran-Cheng/muon-randomized-svd/cifar10/research/airbench94_muon_simple.py f3ac1549c0a496f4 unverified no licence file found · pointer only
The Newton-Muon Optimizer added by Syntology 2026-04 (from id) KellerJordan/modded-nanogpt/records/track_3_optimization/train_gpt_simple.py f21c15e79f628ed4 unverified MIT (permissive)
The Newton-Muon Optimizer added by Syntology 2026-04 (from id) KellerJordan/modded-nanogpt/records/track_1_short/2024-10-10_Muon/train_gpt2.py 5728875012ffc26a unverified MIT (permissive)
The Newton-Muon Optimizer added by Syntology 2026-04 (from id) KellerJordan/modded-nanogpt/records/track_1_short/2024-10-14_ModernArch/train_gpt2.py 063edbb4688922ba unverified MIT (permissive)
The Newton-Muon Optimizer added by Syntology 2026-04 (from id) KellerJordan/modded-nanogpt/records/track_1_short/2024-12-04_ValueEmbed/train_gpt2.py 52f6a37650968e45 unverified MIT (permissive)
Powering Up Zeroth-Order Training via Subspace Gradient Orthogonalization added by Syntology 2026-02 (from id) OPTML-Group/ZO-Muon/llm/LowDimTrainer.py 635c93a80a199840 unverified MIT (permissive)
Adam Improves Muon Adam Improves Muon: Adaptive Moment Estimation with Orthogonalized Momentum added by Syntology 2026-02 (from id) minxin-zhg/namo/src/namo.py c6cbb0406b3d2e05 unverified MIT (permissive)
An Embarrassingly Simple Way to Optimize Orthogonal Matrices at Scale added by Syntology 2026-02 (from id) KellerJordan/cifar10-airbench/airbench94_muon.py 80efc3fcf6a79899 unverified MIT (permissive)
Variance-Adaptive Muon: Accelerating LLM Pretraining with NSR-Modulated and Variance-Scaled Momentum added by Syntology 2026-01 (from id) jingru-lee/Variance-Adaptive-Muon/Suite_A/src/optim/muon.py 04049769e4eed44a ran fingerprinted MIT (permissive)
Variance-Adaptive Muon: Accelerating LLM Pretraining with NSR-Modulated and Variance-Scaled Momentum added by Syntology 2026-01 (from id) jingru-lee/Variance-Adaptive-Muon/Suite_B/muon.py bf190c2d81f45aa9 unverified MIT (permissive)
Variance-Adaptive Muon: Accelerating LLM Pretraining with NSR-Modulated and Variance-Scaled Momentum added by Syntology 2026-01 (from id) jingru-lee/Variance-Adaptive-Muon/Suite_B/muon_nsr.py 901cfa8d98a949c5 unverified MIT (permissive)
Variance-Adaptive Muon: Accelerating LLM Pretraining with NSR-Modulated and Variance-Scaled Momentum added by Syntology 2026-01 (from id) jingru-lee/Variance-Adaptive-Muon/Suite_A/src/optim/muon_vs.py 72f7d6d5b8308b8e unverified MIT (permissive)
FOAM: Blocked State Folding for Memory-Efficient LLM Training added by Syntology 2025-12 (from id) zqOuO/FOAM/foam_torch/muon.py c6cbb0406b3d2e05 unverified no licence file found · pointer only
NorMuon: Making Muon more efficient and scalable added by Syntology 2025-10 (from id) zichongli5/NorMuon/normuon.py b7f41e9dfacd5ee3 ran fingerprinted MIT (permissive)
MERIT: Maximum-normalized Element-wise Ratio for Language Model Large-batch Training added by Syntology 2025-08 (from id) kyleliang919/C-Optim/c_muon.py 88b19bfaad9384f8 ran · our draft was wrong fingerprinted MIT (permissive)
AdaMuon: Adaptive Muon Optimizer 15 Jul 2025 Chongjie-Si/AdaMuon/adamuon.py 5618cdee336eee9f ran · our draft was wrong fingerprinted Apache-2.0 (permissive)
Training Deep Learning Models with Norm-Constrained LMOs 11 Feb 2025 lions-epfl/scion/scion.py ebf4909ddc522784 unverified MIT (permissive)
Training Deep Learning Models with Norm-Constrained LMOs 11 Feb 2025 LunNova/ScionLight-reimpl/scionlight.py 242fd0f7cbb63b14 unverified MIT (permissive)
Cautious Optimizers: Improving Training with One Line of Code 25 Nov 2024 kyleliang919/c-optim/c_muon.py 88b19bfaad9384f8 ran · our draft was wrong fingerprinted MIT (permissive)
Cautious Optimizers: Improving Training with One Line of Code 25 Nov 2024 kyleliang919/c-optim/muon.py c6cbb0406b3d2e05 unverified MIT (permissive)
Large-Scale Contextualised Language Modelling for Norwegian 13 Apr 2021 ltgoslo/NorBERT/norbert4/muon.py c98301effef79e18 unverified CC0-1.0 (permissive)

This site shows no code text; each File cell links to the file on GitHub at the repository's current default branch, which may have changed since the harvest. "Pointer only" means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence cell for the reason. Per-sample records for a paper are on its paper page under "Code Syntology ran".

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections