Prompt optimization algorithms for Large Language Models (LLMs) excel in multi-step reasoning but still lack effective uncertainty estimation. This paper introduces a benchmark dataset to evaluate uncertainty metrics, focusing on Answer, Correctness, Aleatoric, and Epistemic Uncertainty. Through analysis of models like GPT-3.5-Turbo and Meta-Llama-3.1-8B-Instruct, we show that current metrics align more with Answer Uncertainty, which reflects output confidence and diversity, rather than Correctness Uncertainty, highlighting the need for improved metrics that are optimization-objective-aware to better guide prompt optimization. Our code and dataset are available at https://github.com/0Frett/PO-Uncertainty-Benchmarking.
@article{arxiv.2409.10044,
title = {Benchmarking Large Language Model Uncertainty for Prompt Optimization},
author = {Pei-Fu Guo and Yun-Da Tsai and Shou-De Lin},
journal= {arXiv preprint arXiv:2409.10044},
year = {2024}
}