arXiv 2412.057232024-12stat.MLcs.AIcs.CLcs.LG

大语言模型低秩适配器的免训练贝叶斯化

Training-Free Bayesianization for Low-Rank Adapters of Large Language Models

Rutgers University（新泽西罗格斯大学）；University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Red Hat AI Innovation（红帽AI创新）

Haizhou Shi, Yibin Wang, Ligong Han, Huan Zhang, Hao Wang

备注：: Accepted at NeurIPS 2025

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出免训练贝叶斯化（TFB）框架，无需额外训练即可将低秩适配器转为贝叶斯形式，通过搜索最优方差实现不确定性估计与泛化提升。

摘要：

估计大语言模型（LLMs）响应的不确定性仍然是一个关键挑战。尽管最近的贝叶斯方法已证明通过低秩权重更新在量化不确定性方面具有有效性，但它们通常需要复杂的微调或训练后处理流程。在本文中，我们提出了免训练贝叶斯化（TFB），一个简单但具有理论基础的框架，能够在无需额外训练的情况下，高效地将训练好的低秩适配器转换为贝叶斯适配器。TFB系统性地在低秩各向同性高斯分布族内搜索权重后验中最大可接受的方差水平。我们的理论分析表明，在温和条件下，这一搜索过程等价于KL正则化的变分优化，这是变分推断的一种广义形式。通过全面的实验，我们展示了TFB在不确定性估计和泛化方面优于现有方法，同时消除了对复杂贝叶斯化训练流程的需求。代码将在https://github.com/Wang-ML-Lab/bayesian-peft提供。

英文摘要：

Estimating the uncertainty of responses from Large Language Models (LLMs) remains a critical challenge. While recent Bayesian methods have demonstrated effectiveness in quantifying uncertainty through low-rank weight updates, they typically require complex fine-tuning or post-training procedures. In this paper, we propose Training-Free Bayesianization (TFB), a simple yet theoretically grounded framework that efficiently transforms trained low-rank adapters into Bayesian ones without additional training. TFB systematically searches for the maximally acceptable level of variance in the weight posterior, constrained within a family of low-rank isotropic Gaussian distributions. Our theoretical analysis shows that under mild conditions, this search process is equivalent to KL-regularized variational optimization, a generalized form of variational inference. Through comprehensive experiments, we show that TFB achieves superior uncertainty estimation and generalization compared to existing methods while eliminating the need for complex Bayesianization training procedures. Code will be available at https://github.com/Wang-ML-Lab/bayesian-peft.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

AI总结：

摘要：

英文摘要：

相关论文