Evaluating the Capability of Large-scale Language Models on Chinese Grammatical Error Correction Task

Large-scale LLMs has shown remarkable capability in various of NLP tasks and attracted lots of attention recently. However, some studies indicated that LLMs fail to achieve promising result beyond the state-of-the-art models in English grammatical error correction (GEC) tasks. In this report, we aim to explore the how LLMs perform on Chinese grammatical error correction tasks and provide guidance for future work. We conduct experiments with 3 different LLMs of different model scale on 4 Chinese GEC dataset. Our experimental results indicate that the performances of LLMs on automatic evaluation metrics falls short of the previous sota models because of the problem of over-correction. Furthermore, we also discover notable variations in the performance of LLMs when evaluated on different data distributions. Our findings demonstrates that further investigation is required for the application of LLMs on Chinese GEC task.