VAA-CSEC：用于中文语义纠错的投票引导优势分配
VAA-CSEC: Vote-guided Advantage Allocation for Chinese Semantic Error Correction
浏览论文内容
中文总结 AI 辅助
针对中文语义纠错中过度纠正及CoT与自一致性解码交互不清的问题，提出VAA-CSEC多阶段框架，结合CoT蒸馏、SFT、RL和自一致性解码，并引入GLPO对齐训练与推理目标，在CSED-C和NaSGEC-Exam上取得最优F0.5。
中文摘要 AI 辅助
中文语义纠错（CSEC）旨在处理中文文本中的语义错误，这类错误通常比拼写错误和语法错误更为微妙和复杂，但相关研究仍相对不足。现有基于大语言模型（LLM）的方法在此任务中面临两个反复出现的障碍：过度纠正，以及思维链（CoT）推理与自一致性解码之间交互不清晰，导致CoT带来的益处无法可靠地传递给最终纠正结果。我们提出用于CSEC的投票引导优势分配（VAA-CSEC），这是一个多阶段框架，结合了CoT蒸馏、监督微调（SFT）、强化学习（RL）和自一致性解码。在RL阶段，我们设计了一个与CSEC最小编辑原则直接对齐的任务特定奖励函数。我们进一步引入组级相对策略优化（GLPO），该优化根据单个rollout奖励与投票聚合的组奖励之间的差值重新分配GRPO优势，使RL训练目标与推理时使用的自一致性目标对齐。在CSED-C和NaSGEC-Exam上的实验表明，VAA-CSEC在CSED-C上以47.72%的F0.5值超越了所有基于LLM的基线，在所有方法中实现了最高的42.15%召回率，并在NaSGEC-Exam上以41.55%的F0.5值创下了新的最先进水平。
英文摘要
Chinese Semantic Error Correction (CSEC) targets semantic errors in Chinese text, which are typically more subtle and complex than spelling and grammatical errors but remain relatively underexplored. Existing LLM-based approaches face two recurring obstacles in this task: over-correction, and unclear interaction between Chain-of-Thought (CoT) reasoning and self-consistency decoding, such that the benefits brought by CoT cannot be reliably transferred to final corrections. We propose Vote-guided Advantage Allocation for CSEC (VAA-CSEC), a multi-stage framework that combines CoT distillation, Supervised Fine-Tuning (SFT), Reinforcement Learning (RL) and self-consistency decoding. During RL, we design a task-specific reward function that directly aligned with the minimal-editing principle of CSEC. We further introduce Group-Level Relative Policy Optimization (GLPO), which reallocates GRPO advantages according to the margin between individual rollout rewards and the vote-aggregated group reward, aligning the RL training objective with the self-consistency objective used at inference time. Experiments on CSED-C and NaSGEC-Exam show that VAA-CSEC outperforms all LLM-based baselines on CSED-C with an F0.5 of 47.72%, achieves the highest recall of 42.15% among all methods, and establishes a new state of the art of 41.55% F0.5 on NaSGEC-Exam.
发表机构
- Guangdong University of Foreign Studies（广东外语外贸大学）
- National University of Defense Technology（国防科技大学）
机构由 AI 辅助整理，请以论文原文为准。
