AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Learning Rate Scaling across LoRA Ranks and Transfer to Full Finetuning

Nan Chen, Soledad Villar, Soufiane Hayou

ICML 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: LoRA variants and theory · 0 citations (OpenAlex)

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
scale LoRA alpha as r^-1 instead of 1 under Init[A] (ablation)LoRA Init[A] · SFT across five modelsoptimal learning rate vs rankImproved
use Init[B] with alpha=1 instead of Init[A] (ablation)LoRA · SFT, RLVR, diffusionLoRA-to-FFT optimal learning rate alignmentImproved
increase learning rate 2x beyond optimum (ablation)LoRA Init[A] alpha=1 · Qwen2.5-3B-Instruct, RoBERTa-largetraining lossHurt
use Init[B] alpha=1 versus Init[A] alpha=1 for large-LR tolerance (ablation)LoRA · SFTlearning rate stabilityImproved
choose suboptimal learning rate (ablation)LoRA/FFT GRPO · Llama-3.1-8B GSM8kwall-clock training time: 2--4x longerHurt
use LoRA r=256 instead of full finetuning (ablation)Llama-3.1-8B GRPO · Llama-3.1-8B GSM8kpeak GPU memory: FFT over 2x LoRA memoryImproved
alpha in {2, 4} instead of 1 under Init[A] (ablation)LoRA Init[A] · SFTscaling rule r^-1/2No clear effect

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