Dual LoRA: Enhancing LoRA with Magnitude and Direction Updates

ACL ARR 2026 January Submission257 Authors

Keywords: LoRA, parameter-efficient fine-tuning, LLM

Abstract: Low-rank adaptation (LoRA) is one of the most popular methods among parameter-efficient fine-tuning (PEFT) methods to adapt pre-trained large language models (LLMs) to specific downstream tasks. However, the model trained based on LoRA often has an unsatisfactory performance due to its low-rank assumption. In this paper, we propose a novel method called Dual LoRA to improve the performance by incorporating an inductive bias into the original LoRA. Specifically, we separate low-rank matrices into two groups: the magnitude group to control whether or not and how far we should update a parameter and the direction group to decide whether this parameter should move forward or backward, to better simulate the parameter updating process of the full fine-tuning based on gradient-based optimization algorithms. We show that this can be simply achieved by adding a ReLU function to the magnitude group and a sign function to the direction group. We conduct several experiments over a wide range of NLP tasks, including natural language understanding (NLU), generation (NLG), and commonsense reasoning datasets on LLaMA-1/2/3, RoBERTa, DeBERTa, and GPT2 as baseline models. Results show that we consistently outperform LoRA and its SOTA variants with the same number of trainable parameters.

Paper Type: Long

Research Area: Low-resource Methods for NLP

Research Area Keywords: Efficient/Low-Resource Methods for NLP

Contribution Types: Approaches to low-resource settings

Languages Studied: English

Submission Number: 257