In this paper, we introduce InfiAgent-DABench, the first benchmark specifically designed to evaluate LLM-based agents on data analysis tasks. These tasks require agents to end-to-end solving complex tasks by interacting with an execution environment. This benchmark contains DAEval, a dataset consisting of 257 data analysis questions derived from 52 CSV files, and an agent framework which incorporates LLMs to serve as data analysis agents for both serving and evaluation. Since data analysis questions are often open-ended and hard to evaluate without human supervision, we adopt a format-prompting technique to convert each question into a closed-form format so that they can be automatically evaluated. Our extensive benchmarking of 34 LLMs uncovers the current challenges encountered in data analysis tasks. In addition, building on top of our agent framework, we develop a specialized agent, DAAgent, which surpasses GPT-3.5 by 3.9% on DABench. Evaluation datasets and toolkits for InfiAgent-DABench are released at https://github.com/InfiAgent/InfiAgent .
@article{arxiv.2401.05507,
title = {InfiAgent-DABench: Evaluating Agents on Data Analysis Tasks},
author = {Xueyu Hu and Ziyu Zhao and Shuang Wei and Ziwei Chai and Qianli Ma and Guoyin Wang and Xuwu Wang and Jing Su and Jingjing Xu and Ming Zhu and Yao Cheng and Jianbo Yuan and Jiwei Li and Kun Kuang and Yang Yang and Hongxia Yang and Fei Wu},
journal= {arXiv preprint arXiv:2401.05507},
year = {2024}
}