arXiv cs.CL· Ryszard Staruch, Filip Grali\'nski·· 11 小时前AI 评分27
通过重复惩罚在推理阶段调节 LLM 语法纠错的精确率-召回率权衡
Precision--Recall Trade-off on the Inference Level for LLMs in Grammatical Error Correction
AI 导读
研究发现，重复惩罚参数可在不修改模型架构、训练流程或提示词的情况下，于推理阶段可控地调节 LLM 语法纠错（GEC）的精确率-召回率权衡。在标准 GEC 基准上的实验显示，这种基于重复惩罚的解码方式在概念上与序列到序列 GEC 模型此前提出的 Align-and-Predict Decoding（APD）相关，且实现简单，适用于暴露推理阶段解码惩罚参数的现有 LLM。
正文 · 原文
Abstract:Large language models (LLMs) have recently been adopted for Grammatical Error Correction (GEC), achieving strong performance via both full fine-tuning and prompt-based approaches. As probabilistic sequence models, LLMs expose inference-time parameters that directly modify token generation behavior; however, their potential for controlling GEC-specific trade-offs has not been systematically studied. In this work, we investigate decoding strategies for grammatical error correction. We show that the repetition penalty parameter enables controllable adjustment of the precision--recall trade-off at inference time, without modifying the model architecture, training procedure, or prompt. Through experiments on standard GEC benchmarks, we demonstrate that repetition-penalty-based decoding enables inference-time control of the precision--recall balance, conceptually related to methods previously proposed for sequence-to-sequence GEC models, such as Align-and-Predict Decoding (APD), while remaining simple and broadly applicable to existing LLMs that expose inference-time decoding penalties.
| Comments: | 17 pages, 2 figures, 13 tables |
| Subjects: | Computation and Language (cs.CL) |
| ACM classes: | I.2.7 |
| Cite as: | arXiv:2610.41698 [cs.CL] |
| (or arXiv:2610.41698v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41698 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ryszard Staruch [view email]
[v1]
Fri, 30 Oct 2026 13:04:17 UTC (94 KB)
来源：arXiv cs.CL · arxiv.org