paper-with-me

Papers

Language Model Uncertainty Quantification with Attention Chain

2025-03-24 · Yinghao Li, Rushi Qiang, Lama Moukheiber, Chao Zhang

Accurately quantifying a large language model's (LLM) predictive uncertainty is crucial for judging the reliability of its answers. While most existing research focuses on short, directly answerable questions with closed-form outputs (e.g., multiple-choice), involving intermediate reasoning steps in LLM responses is increasingly important. This added complexity complicates uncertainty quantification (UQ) because the probabilities assigned to answer tokens are conditioned on a vast space of preceding reasoning tokens. Direct marginalization is infeasible, and the dependency inflates probability estimates, causing overconfidence in UQ. To address this, we propose UQAC, an efficient method that narrows the reasoning space to a tractable size for marginalization. UQAC iteratively constructs an "attention chain" of tokens deemed "semantically crucial" to the final answer via a backtracking procedure. Starting from the answer tokens, it uses attention weights to identify the most influential predecessors, then iterates this process until reaching the input tokens. Similarity filtering and probability thresholding further refine the resulting chain, allowing us to approximate the marginal probabilities of the answer tokens, which serve as the LLM's confidence. We validate UQAC on multiple reasoning benchmarks with advanced open-source LLMs, demonstrating that it consistently delivers reliable UQ estimates with high computational efficiency.

📄 PDF Abstract BibTeX arXiv:2503.19168

Code (1)

Yinghao-Li/UQAC 공식 구현

Tasks

Computational EfficiencyLanguage ModelingLanguage ModellingmodelMultiple-choiceUncertainty Quantification

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

Simple Yet Effective: An Information-Theoretic Approach to Multi-LLM Uncertainty Quantification

2025-07-09 · Maya Kruse, Majid Afshar, Saksham Khatwani, Anoop Mayampurath 외 arxiv

Large language models (LLMs) often behave inconsistently across inputs, indicating uncertainty and motivating the need for its quantification in high-stakes settings. Prior work on calibration and uncertainty quantificat…

Uncertainty quantification in non-rigid image registration via stochastic gradient Markov chain Monte Carlo

2021-10-25 · Daniel Grzech, Mohammad Farid Azampour, Huaqi Qiu, Ben Glocker 외

We develop a new Bayesian model for non-rigid registration of three-dimensional medical images, with a focus on uncertainty quantification. Probabilistic registration of large images with calibrated uncertainty estimates…

Image RegistrationUncertainty QuantificationVariational Inference

Inv-Entropy: A Fully Probabilistic Framework for Uncertainty Quantification in Language Models

2025-06-11 · Haoyi Song, Ruihan Ji, Naichen Shi, Fan Lai 외

Large language models (LLMs) have transformed natural language processing, but their reliable deployment requires effective uncertainty quantification (UQ). Existing UQ methods are often heuristic and lack a probabilisti…

DiversitySemantic SimilaritySemantic Textual SimilarityUncertainty Quantification

Uncertainty-Aware Attention Heads: Efficient Unsupervised Uncertainty Quantification for LLMs

2025-05-26 · Artem Vazhentsev, Lyudmila Rvanova, Gleb Kuzmin, Ekaterina Fadeeva 외

Large language models (LLMs) exhibit impressive fluency, but often produce critical errors known as "hallucinations". Uncertainty quantification (UQ) methods are a promising tool for coping with this fundamental shortcom…

HallucinationQuestion AnsweringUncertainty Quantification

Bayesian graph convolutional neural networks via tempered MCMC

2021-04-17 · Rohitash Chandra, Ayush Bhagat, Manavendra Maharana, Pavel N. Krivitsky

Deep learning models, such as convolutional neural networks, have long been applied to image and multi-media tasks, particularly those with structured data. More recently, there has been more attention to unstructured da…

Bayesian InferenceDeep LearningUncertainty Quantification