Saved in:
Bibliographic Details
Main Authors: Cao, Kefan, Wu, Shuaicheng
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.06100
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908782988099584
author Cao, Kefan
Wu, Shuaicheng
author_facet Cao, Kefan
Wu, Shuaicheng
contents Large language models (LLMs) suffer from catastrophic forgetting in sequential multi-task learning. Existing parameter regularization methods (e.g., O-LoRA, N-LoRA) mitigate interference via low-rank subspace orthogonality, but additive updates distort the intrinsic geometry of model parameters. We propose \textbf{OLieRA}, a Lie group based fine-tuning framework that preserves parameter geometry through multiplicative updates while enforcing orthogonality across task subspaces. OLieRA achieves state-of-the-art performance on the Standard CL benchmark and remains highly competitive under large task sequences. It further inherits the replay-free and task-ID free inference properties of O-LoRA, establishing a principled paradigm for continual learning in LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2509_06100
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Orthogonal Low-rank Adaptation in Lie Groups for Continual Learning of Large Language Models
Cao, Kefan
Wu, Shuaicheng
Computation and Language
Large language models (LLMs) suffer from catastrophic forgetting in sequential multi-task learning. Existing parameter regularization methods (e.g., O-LoRA, N-LoRA) mitigate interference via low-rank subspace orthogonality, but additive updates distort the intrinsic geometry of model parameters. We propose \textbf{OLieRA}, a Lie group based fine-tuning framework that preserves parameter geometry through multiplicative updates while enforcing orthogonality across task subspaces. OLieRA achieves state-of-the-art performance on the Standard CL benchmark and remains highly competitive under large task sequences. It further inherits the replay-free and task-ID free inference properties of O-LoRA, establishing a principled paradigm for continual learning in LLMs.
title Orthogonal Low-rank Adaptation in Lie Groups for Continual Learning of Large Language Models
topic Computation and Language
url https://arxiv.org/abs/2509.06100