CEC-Zero: Chinese Error Correction Solution Based on LLM

TL;DR

This paper introduces CEC-Zero, a reinforcement learning framework that enables large language models to self-correct Chinese text errors autonomously, improving accuracy and generalization without relying on annotated data.

Contribution

The paper presents a novel RL-based method that allows LLMs to self-correct Chinese spelling errors, reducing dependence on external supervision and enhancing cross-domain robustness.

Findings

RL-enhanced LLMs achieve industry-viable accuracy

Method improves cross-domain generalization

Eliminates need for annotated data or auxiliary models

Abstract

Recent advancements in large language models (LLMs) demonstrate exceptional Chinese text processing capabilities, particularly in Chinese Spelling Correction (CSC). While LLMs outperform traditional BERT-based models in accuracy and robustness, challenges persist in reliability and generalization. This paper proposes CEC-Zero, a novel reinforcement learning (RL) framework enabling LLMs to self-correct through autonomous error strategy learning without external supervision. By integrating RL with LLMs' generative power, the method eliminates dependency on annotated data or auxiliary models. Experiments reveal RL-enhanced LLMs achieve industry-viable accuracy and superior cross-domain generalization, offering a scalable solution for reliability optimization in Chinese NLP applications. This breakthrough facilitates LLM deployment in practical Chinese text correction scenarios while establishing a new paradigm for self-improving language models.