Measuring Taiwanese Mandarin Language Understanding

TL;DR

This paper introduces TMLU, a comprehensive benchmark for evaluating Taiwanese Mandarin language understanding in LLMs, highlighting performance gaps and fostering development of localized models.

Contribution

It presents TMLU, a new evaluation suite for Taiwanese Mandarin LLMs, including diverse subjects, reasoning tasks, and a baseline analysis of 24 models.

Findings

Chinese open-weight models perform worse than multilingual proprietary models.

Taiwanese Mandarin models lag behind Simplified Chinese counterparts.

Significant room for improvement in Taiwanese Mandarin LLMs.

Abstract

The evaluation of large language models (LLMs) has drawn substantial attention in the field recently. This work focuses on evaluating LLMs in a Chinese context, specifically, for Traditional Chinese which has been largely underrepresented in existing benchmarks. We present TMLU, a holistic evaluation suit tailored for assessing the advanced knowledge and reasoning capability in LLMs, under the context of Taiwanese Mandarin. TMLU consists of an array of 37 subjects across social science, STEM, humanities, Taiwan-specific content, and others, ranging from middle school to professional levels. In addition, we curate chain-of-thought-like few-shot explanations for each subject to facilitate the evaluation of complex reasoning skills. To establish a comprehensive baseline, we conduct extensive experiments and analysis on 24 advanced LLMs. The results suggest that Chinese open-weight models demonstrate inferior performance comparing to multilingual proprietary ones, and open-weight models tailored for Taiwanese Mandarin lag behind the Simplified-Chinese counterparts. The findings indicate great headrooms for improvement, and emphasize the goal of TMLU to foster the development of localized Taiwanese-Mandarin LLMs. We release the benchmark and evaluation scripts for the community to promote future research.