Saved in:
Bibliographic Details
Main Authors: Zhang, Kaituo, Hu, Mingzhi, Le, Hoang Anh Duy, Torsha, Fariha Kabir, Jiang, Zhimeng, Bui, Minh Khai, Chang, Chia-Yuan, Chuang, Yu-Neng, Xiong, Zhen, Lin, Ying, Wang, Guanchu, Zou, Na
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.17717
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908790116319232
author Zhang, Kaituo
Hu, Mingzhi
Le, Hoang Anh Duy
Torsha, Fariha Kabir
Jiang, Zhimeng
Bui, Minh Khai
Chang, Chia-Yuan
Chuang, Yu-Neng
Xiong, Zhen
Lin, Ying
Wang, Guanchu
Zou, Na
author_facet Zhang, Kaituo
Hu, Mingzhi
Le, Hoang Anh Duy
Torsha, Fariha Kabir
Jiang, Zhimeng
Bui, Minh Khai
Chang, Chia-Yuan
Chuang, Yu-Neng
Xiong, Zhen
Lin, Ying
Wang, Guanchu
Zou, Na
contents Large Language Models (LLMs) have emerged as powerful tools for generating data across various modalities. By transforming data from a scarce resource into a controllable asset, LLMs mitigate the bottlenecks imposed by the acquisition costs of real-world data for model training, evaluation, and system iteration. However, ensuring the high quality of LLM-generated synthetic data remains a critical challenge. Existing research primarily focuses on generation methodologies, with limited direct attention to the quality of the resulting data. Furthermore, most studies are restricted to single modalities, lacking a unified perspective across different data types. To bridge this gap, we propose the \textbf{LLM Data Auditor framework}. In this framework, we first describe how LLMs are utilized to generate data across six distinct modalities. More importantly, we systematically categorize intrinsic metrics for evaluating synthetic data from two dimensions: quality and trustworthiness. This approach shifts the focus from extrinsic evaluation, which relies on downstream task performance, to the inherent properties of the data itself. Using this evaluation system, we analyze the experimental evaluations of representative generation methods for each modality and identify substantial deficiencies in current evaluation practices. Based on these findings, we offer concrete recommendations for the community to improve the evaluation of data generation. Finally, the framework outlines methodologies for the practical application of synthetic data across different modalities.
format Preprint
id arxiv_https___arxiv_org_abs_2601_17717
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle The LLM Data Auditor: A Metric-oriented Survey on Quality and Trustworthiness in Evaluating Synthetic Data
Zhang, Kaituo
Hu, Mingzhi
Le, Hoang Anh Duy
Torsha, Fariha Kabir
Jiang, Zhimeng
Bui, Minh Khai
Chang, Chia-Yuan
Chuang, Yu-Neng
Xiong, Zhen
Lin, Ying
Wang, Guanchu
Zou, Na
Artificial Intelligence
Machine Learning
Large Language Models (LLMs) have emerged as powerful tools for generating data across various modalities. By transforming data from a scarce resource into a controllable asset, LLMs mitigate the bottlenecks imposed by the acquisition costs of real-world data for model training, evaluation, and system iteration. However, ensuring the high quality of LLM-generated synthetic data remains a critical challenge. Existing research primarily focuses on generation methodologies, with limited direct attention to the quality of the resulting data. Furthermore, most studies are restricted to single modalities, lacking a unified perspective across different data types. To bridge this gap, we propose the \textbf{LLM Data Auditor framework}. In this framework, we first describe how LLMs are utilized to generate data across six distinct modalities. More importantly, we systematically categorize intrinsic metrics for evaluating synthetic data from two dimensions: quality and trustworthiness. This approach shifts the focus from extrinsic evaluation, which relies on downstream task performance, to the inherent properties of the data itself. Using this evaluation system, we analyze the experimental evaluations of representative generation methods for each modality and identify substantial deficiencies in current evaluation practices. Based on these findings, we offer concrete recommendations for the community to improve the evaluation of data generation. Finally, the framework outlines methodologies for the practical application of synthetic data across different modalities.
title The LLM Data Auditor: A Metric-oriented Survey on Quality and Trustworthiness in Evaluating Synthetic Data
topic Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2601.17717