Papers
arxiv:2608.31036

Normalized Low-Rank Adaptation

Published on Aug 31
· Submitted by
J.L
on Sep 1
Authors:
,
,
,
,

Abstract

Normalized Low-Rank Adaptation stabilizes LoRA training by normalizing down-projection matrices, accelerating convergence and improving performance without extra parameters or inference cost.

While low-rank adaptation (LoRA) is widely used for parameter-efficient model adaptation, how to regularize its training dynamics for stable and effective optimization remains underexplored. Because LoRA initializes the up-projection to zero, its early optimization dynamics are largely governed by the down-projection. Building on this observation, we introduce Normalized Low-Rank Adaptation (NoRA), a simple yet effective method that normalizes the down-projection matrices during training. We further show that the same normalization can be applied only at initialization, improving standard LoRA without requiring repeated normalization throughout training. Across pretraining, supervised finetuning, and reinforcement learning, NoRA consistently accelerates convergence, improves performance and training stability, and mitigates catastrophic forgetting. These benefits require neither additional trainable parameters nor inference-time computation, making NoRA a simple and broadly applicable enhancement to LoRA.

Community

Paper submitter

to be better

I think this paper has in common with idea in https://arxiv.org/abs/2510.01938

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2608.31036 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2608.31036 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2608.31036 in a Space README.md to link it from this page.

Collections including this paper 5