C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for Foundation Models

TL;DR

C-Eval is a comprehensive Chinese evaluation suite with multi-level, multi-discipline questions designed to assess the reasoning and knowledge abilities of foundation models, revealing current limitations of state-of-the-art LLMs like GPT-4.

Contribution

This paper introduces C-Eval, the first extensive Chinese benchmark covering multiple disciplines and difficulty levels to evaluate advanced LLM capabilities.

Findings

GPT-4 achieves over 60% accuracy on C-Eval

Most LLMs still have significant room for improvement

C-Eval reveals strengths and weaknesses of foundation models

Abstract

New NLP benchmarks are urgently needed to align with the rapid development of large language models (LLMs). We present C-Eval, the first comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. C-Eval comprises multiple-choice questions across four difficulty levels: middle school, high school, college, and professional. The questions span 52 diverse disciplines, ranging from humanities to science and engineering. C-Eval is accompanied by C-Eval Hard, a subset of very challenging subjects in C-Eval that requires advanced reasoning abilities to solve. We conduct a comprehensive evaluation of the most advanced LLMs on C-Eval, including both English- and Chinese-oriented models. Results indicate that only GPT-4 could achieve an average accuracy of over 60%, suggesting that there is still significant room for improvement for current LLMs. We anticipate C-Eval will help analyze important strengths and shortcomings of foundation models, and foster their development and growth for Chinese users.