paper-with-me

Papers

Efficient Uncertainty Estimation via Distillation of Bayesian Large Language Models

2025-05-16 · Harshil Vejendla, Haizhou Shi, Yibin Wang, Tunyu Zhang, huan zhang, Hao Wang

Recent advances in uncertainty estimation for Large Language Models (LLMs) during downstream adaptation have addressed key challenges of reliability and simplicity. However, existing Bayesian methods typically require multiple sampling iterations during inference, creating significant efficiency issues that limit practical deployment. In this paper, we investigate the possibility of eliminating the need for test-time sampling for LLM uncertainty estimation. Specifically, when given an off-the-shelf Bayesian LLM, we distill its aligned confidence into a non-Bayesian student LLM by minimizing the divergence between their predictive distributions. Unlike typical calibration methods, our distillation is carried out solely on the training dataset without the need of an additional validation dataset. This simple yet effective approach achieves N-times more efficient uncertainty estimation during testing, where N is the number of samples traditionally required by Bayesian LLMs. Our extensive experiments demonstrate that uncertainty estimation capabilities on training data can successfully generalize to unseen test data through our distillation technique, consistently producing results comparable to (or even better than) state-of-the-art Bayesian LLMs.

📄 PDF Abstract BibTeX arXiv:2505.11731

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Iterative Distillation for Better Uncertainty Estimates in Multitask Emotion Recognition

2021-07-21 · Didan Deng, Liang Wu, Bertram E. Shi

When recognizing emotions, subtle nuances in displays of emotion generate ambiguity or uncertainty in emotion perception. Emotion uncertainty has been previously interpreted as inter-rater disagreement among multiple ann…

Emotion Recognition

Uncertainty-Aware Knowledge Distillation for Multimodal Large Language Models

2026-03-22 · Jingchen Sun, Shaobo Han, Deep Patel, Wataru Kohno 외 arxiv

Knowledge distillation establishes a learning paradigm that leverages both data supervision and teacher guidance. However, determining the optimal balance between learning from data and learning from the teacher is chall…

Knowledge Distillation

Multi-Teacher Knowledge Distillation via Teacher-Informed Mixture Priors

2026-05-27 · Luyang Fang, Yongkai Chen, Jiazhang Cai, Ping Ma 외 arxiv

Knowledge distillation is a powerful method for model compression, enabling the efficient deployment of complex deep learning models (teachers), including large language models. However, its underlying statistical mechan…

Knowledge DistillationImage ClassificationBayesian InferenceModel Compression

Bayesian Mixture of Experts For Large Language Models

2025-11-12 · Maryam Dialameh, Hossein Rajabzadeh, Weiwei Zhang, Walid Ahmed 외 arxiv

We present Bayesian Mixture of Experts (Bayesian-MoE), a post-hoc uncertainty estimation framework for fine-tuned large language models (LLMs) based on Mixture-of-Experts architectures. Our method applies a structured La…

Bayesian Inference

BLoB: Bayesian Low-Rank Adaptation by Backpropagation for Large Language Models

2024-06-17 · Yibin Wang, Haizhou Shi, Ligong Han, Dimitris Metaxas 외

Large Language Models (LLMs) often suffer from overconfidence during inference, particularly when adapted to downstream domain-specific tasks with limited data. Previous work addresses this issue by employing approximate…