Decoding Uncertainty: The Impact of Decoding Strategies for Uncertainty Estimation in Large Language Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911165928439808 |
|---|---|
| author | Hashimoto, Wataru Kamigaito, Hidetaka Watanabe, Taro |
| author_facet | Hashimoto, Wataru Kamigaito, Hidetaka Watanabe, Taro |
| contents | Decoding strategies manipulate the probability distribution underlying the output of a language model and can therefore affect both generation quality and its uncertainty. In this study, we investigate the impact of decoding strategies on uncertainty estimation in Large Language Models (LLMs). Our experiments show that Contrastive Search, which mitigates repetition, yields better uncertainty estimates on average across a range of preference-aligned LLMs. In contrast, the benefits of these strategies sometimes diverge when the model is only post-trained with supervised fine-tuning, i.e. without explicit alignment. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_16696 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Decoding Uncertainty: The Impact of Decoding Strategies for Uncertainty Estimation in Large Language Models Hashimoto, Wataru Kamigaito, Hidetaka Watanabe, Taro Computation and Language Machine Learning Decoding strategies manipulate the probability distribution underlying the output of a language model and can therefore affect both generation quality and its uncertainty. In this study, we investigate the impact of decoding strategies on uncertainty estimation in Large Language Models (LLMs). Our experiments show that Contrastive Search, which mitigates repetition, yields better uncertainty estimates on average across a range of preference-aligned LLMs. In contrast, the benefits of these strategies sometimes diverge when the model is only post-trained with supervised fine-tuning, i.e. without explicit alignment. |
| title | Decoding Uncertainty: The Impact of Decoding Strategies for Uncertainty Estimation in Large Language Models |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2509.16696 |