AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

LoRA meets Riemannion: Muon Optimizer for Parametrization-independent Low-Rank Adapters

Vladimir Bogachev, Vladimir Aletov, Alexander Molozhavenko, Denis Bobkov, Vera Soboleva, Aibek Alanov, Maxim Rakhuba

ICLR 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: LoRA variants and theory

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace Adam-LoRA with Riemannion optimizerLoRA rank 16 · Commonsense reasoning (8 tasks avg), Llama 3-8Baccuracy (%): 87.1 -> 88.1Improved
use Riemannion instead of Muon applied per LoRA factorLoRA rank 16 · Commonsense reasoning (8 tasks avg), Llama 3-8Baccuracy (%): 83.0 -> 88.1Improved
use Riemannion instead of RPrecAdamW (Riemannian-aware Adam variant)LoRA rank 16 · Commonsense reasoning (8 tasks avg), Llama 3-8Baccuracy (%): 86.8 -> 88.1Improved
apply Muon directly to LoRA factorsLoRA Adam baseline · Commonsense reasoning (8 tasks avg), Llama 3-8Baccuracy (%): 87.1 -> 83.0Hurt
add locally optimal initialization (LOI) to LoRA (ablation)LoRA · Commonsense reasoning (8 tasks avg), Llama 3.2-1Baccuracy (%): 67.9 -> 70.1Improved
add LOI to Riemannian SGD-type LoRA (RiemannLoRA) (ablation)RiemannLoRA · Commonsense reasoning (8 tasks avg), Llama 3.2-1Baccuracy (%): 69.6 -> 73.4Improved
use Riemannian optimization instead of Euclidean LoRA (no LOI) (ablation)LoRA · Commonsense reasoning (8 tasks avg), Llama 3.2-1Baccuracy (%): 67.9 -> 69.6Improved
Riemannion vs Adam LoRA on individual tasksLoRA rank 16 · BoolQ, Llama 3-8Baccuracy (%): 74.8 -> 75.7No clear effect
solve tangent-space LMO exactly or by alternating projection instead of one projection step (ablation)Riemannion · convergenceconvergenceNo clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy