Benchmarking Large Language Model Uncertainty for Prompt Optimization
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915080057126912 |
|---|---|
| author | Guo, Pei-Fu Tsai, Yun-Da Lin, Shou-De |
| author_facet | Guo, Pei-Fu Tsai, Yun-Da Lin, Shou-De |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_10044 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Benchmarking Large Language Model Uncertainty for Prompt Optimization Guo, Pei-Fu Tsai, Yun-Da Lin, Shou-De Machine Learning Computation and Language 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. |
| title | Benchmarking Large Language Model Uncertainty for Prompt Optimization |
| topic | Machine Learning Computation and Language |
| url | https://arxiv.org/abs/2409.10044 |