Revisiting Uncertainty Quantification Evaluation in Language Models: Spurious Interactions with Response Length Bias Results
원문 초록
Uncertainty Quantification (UQ) in Language Models (LMs) is crucial for improving their safety and reliability. Evaluations often use performance metrics like AUROC to assess how well UQ methods (e.g., negative sequence probabilities) correlate with task correctness functions (e.g., ROUGE-L). In this paper, we show that commonly used correctness functions bias UQ evaluations by inflating the performance of certain UQ methods. We evaluate 7 correctness functions -- from lexical-based and embedding-based metrics to LLM-as-a-judge approaches -- across 4 datasets x 4 models x 6 UQ methods. Our analysis reveals that length biases in the errors of these correctness functions distort UQ assessments by interacting with length biases in UQ methods. We identify LLM-as-a-judge approaches as among the least length-biased choices and hence a potential solution to mitigate these biases.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Statistical Machine Learning for Astronomy -- A Textbook
Interpretable Bayesian Tensor Network Kernel Machines with Automatic Rank and Feature Selection
Estimating prevalence with precision and accuracy
UQLM: A Python Package for Uncertainty Quantification in Large Language Models
Scalable Bayesian Low-Rank Adaptation of Large Language Models via Stochastic Variational Subspace Inference
Consensus-Driven Uncertainty for Robotic Grasping based on RGB Perception
토론
이 논문과 연결된 포럼 글이 아직 없습니다.