The Effects of Data Augmentation on Confidence Estimation for LLMs

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wang, Rui, Zhu, Renyu, Lin, Minmin, Wu, Runze, Lv, Tangjie, Fan, Changjie, Wang, Haobo
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913890765373440
author Wang, Rui
Zhu, Renyu
Lin, Minmin
Wu, Runze
Lv, Tangjie
Fan, Changjie
Wang, Haobo
author_facet Wang, Rui
Zhu, Renyu
Lin, Minmin
Wu, Runze
Lv, Tangjie
Fan, Changjie
Wang, Haobo
contents Confidence estimation is crucial for reflecting the reliability of large language models (LLMs), particularly in the widely used closed-source models. Utilizing data augmentation for confidence estimation is viable, but discussions focus on specific augmentation techniques, limiting its potential. We study the impact of different data augmentation methods on confidence estimation. Our findings indicate that data augmentation strategies can achieve better performance and mitigate the impact of overconfidence. We investigate the influential factors related to this and discover that, while preserving semantic information, greater data diversity enhances the effectiveness of augmentation. Furthermore, the impact of different augmentation strategies varies across different range of application. Considering parameter transferability and usability, the random combination of augmentations is a promising choice.
format Preprint
id arxiv_https___arxiv_org_abs_2506_11046
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle The Effects of Data Augmentation on Confidence Estimation for LLMs
Wang, Rui
Zhu, Renyu
Lin, Minmin
Wu, Runze
Lv, Tangjie
Fan, Changjie
Wang, Haobo
Machine Learning
Confidence estimation is crucial for reflecting the reliability of large language models (LLMs), particularly in the widely used closed-source models. Utilizing data augmentation for confidence estimation is viable, but discussions focus on specific augmentation techniques, limiting its potential. We study the impact of different data augmentation methods on confidence estimation. Our findings indicate that data augmentation strategies can achieve better performance and mitigate the impact of overconfidence. We investigate the influential factors related to this and discover that, while preserving semantic information, greater data diversity enhances the effectiveness of augmentation. Furthermore, the impact of different augmentation strategies varies across different range of application. Considering parameter transferability and usability, the random combination of augmentations is a promising choice.
title The Effects of Data Augmentation on Confidence Estimation for LLMs
topic Machine Learning
url https://arxiv.org/abs/2506.11046