BEFT: Bias-Efficient Fine-Tuning of Language Models in Low-Data Regimes

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Huang, Baichuan, Balashankar, Ananth, Aminifar, Amir
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913043839975424
author Huang, Baichuan
Balashankar, Ananth
Aminifar, Amir
author_facet Huang, Baichuan
Balashankar, Ananth
Aminifar, Amir
contents Fine-tuning the bias terms of large language models (LLMs) has the potential to achieve unprecedented parameter efficiency while maintaining competitive performance, particularly in low-data regimes. However, the link between fine-tuning different bias terms (i.e., $\boldsymbol{b}_q$, $\boldsymbol{b}_k$, and $\boldsymbol{b}_v$ in the query, key, or value projections) and downstream performance remains largely unclear to date. In this paper, we investigate the link between fine-tuning $\boldsymbol{b}_q$, $\boldsymbol{b}_k$, and $\boldsymbol{b}_v$ with the performance of the downstream task. Our key finding is that directly fine-tuning $\boldsymbol{b}_v$ generally leads to higher downstream performance in low-data regimes, in comparison to $\boldsymbol{b}_q$ and $\boldsymbol{b}_k$. We extensively evaluate this unique property across a wide range of LLMs spanning encoder-only and decoder-only architectures up to 6.7B parameters (including bias-free LLMs). Our results provide strong evidence for the effectiveness of directly fine-tuning $\boldsymbol{b}_v$ across various downstream tasks. The implementation code is available at https://github.com/whubaichuan/BEFT.
format Preprint
id arxiv_https___arxiv_org_abs_2509_15974
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BEFT: Bias-Efficient Fine-Tuning of Language Models in Low-Data Regimes
Huang, Baichuan
Balashankar, Ananth
Aminifar, Amir
Computation and Language
Artificial Intelligence
Machine Learning
Fine-tuning the bias terms of large language models (LLMs) has the potential to achieve unprecedented parameter efficiency while maintaining competitive performance, particularly in low-data regimes. However, the link between fine-tuning different bias terms (i.e., $\boldsymbol{b}_q$, $\boldsymbol{b}_k$, and $\boldsymbol{b}_v$ in the query, key, or value projections) and downstream performance remains largely unclear to date. In this paper, we investigate the link between fine-tuning $\boldsymbol{b}_q$, $\boldsymbol{b}_k$, and $\boldsymbol{b}_v$ with the performance of the downstream task. Our key finding is that directly fine-tuning $\boldsymbol{b}_v$ generally leads to higher downstream performance in low-data regimes, in comparison to $\boldsymbol{b}_q$ and $\boldsymbol{b}_k$. We extensively evaluate this unique property across a wide range of LLMs spanning encoder-only and decoder-only architectures up to 6.7B parameters (including bias-free LLMs). Our results provide strong evidence for the effectiveness of directly fine-tuning $\boldsymbol{b}_v$ across various downstream tasks. The implementation code is available at https://github.com/whubaichuan/BEFT.
title BEFT: Bias-Efficient Fine-Tuning of Language Models in Low-Data Regimes
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2509.15974