Edit-level Majority Voting Mitigates Over-Correction in LLM-based Grammatical Error Correction
Not on the map. Why?
Abstract (arXiv)
Grammatical error correction using large language models often suffers from the over-correction issue. To mitigate this, we propose a training-free inference method that performs edit-level majority voting over multiple candidates generated by a single model, without requiring model modifications or additional training. Across nine benchmarks covering English, Czech, German, Ukrainian, Korean, Hindi, and Romanian, the proposed method outperforms both greedy and MBR decoding in most cases. Moreover, it yields stable correction quality regardless of the instruction prompts used. We release two repository supporting GEC datasets loading and LLM inference.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 31 of its 62 references that are in the map. First posted on arXiv 13 May 2026.
Builds on
31 of 62 references.
References · 31
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
- 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
- 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
- 2019The Curious Case of Neural Text Degenerationcited by 2,881
- 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,928
- 2021DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharingcited by 1,219
- 2020Stanza: A Python Natural Language Processing Toolkit for Many Human Languagescited by 719
- 2020GECToR -- Grammatical Error Correction: Tag, Not Rewritecited by 177
- 2020Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural Machine Translationcited by 137
- 2021A Simple Recipe for Multilingual Grammatical Error Correctioncited by 108
- 2017JFLEG: A Fluency Corpus and Benchmark for Grammatical Error Correctioncited by 96
- 2019Parallel Iterative Edit Models for Local Sequence Transductioncited by 81
19 more
- 2019Improving Grammatical Error Correction via Pre-Training a Copy-Augmented Architecture with Unlabeled Datacited by 50
- 2023Exploring Effectiveness of GPT-3 in Grammatical Error Correction: A Study on Performance and Controllability in Prompt-Based Methodscited by 37
- 2022SynGEC: Syntax-Enhanced Grammatical Error Correction with a Tailored GEC-Oriented Parsercited by 32
- 2019Grammatical Error Correction in Low-Resource Scenarioscited by 28
- 2016GLEU Without Tuningcited by 24
- 2024Prompting open-source and commercial language models for grammatical error correction of English learner textcited by 24
- 2022Revisiting Grammatical Error Correction Evaluation and Beyondcited by 18
- 2024GPT-3.5 for Grammatical Error Correctioncited by 18
- 2020Grammatical Error Correction in Low Error Density Domains: A New Benchmark and Analysescited by 16
- 2022Towards standardizing Korean Grammatical Error Correction: Datasets and Annotationcited by 13
- 2024mbrs: A Library for Minimum Bayes Risk Decodingcited by 13
- 2025gec-metrics: A Unified Library for Grammatical Error Correction Evaluationcited by 9
- 2021System Combination for Grammatical Error Correction Based on Integer Programmingcited by 7
- 2024Toward the Evaluation of Large Language Models Considering Score Variance across Instruction Templatescited by 7
- 2025Rethinking Evaluation Metrics for Grammatical Error Correction: Why Use a Different Evaluation Process than Human?cited by 7
- 2025Adapting LLMs for Minimal-edit Grammatical Error Correctioncited by 5
- 2023Minimum Bayes' Risk Decoding for System Combination of Grammatical Error Correction Systemscited by 4
- 2025Multilingual Grammatical Error Annotation: Combining Language-Agnostic Framework with Language-Specific Flexibilitycited by 2
- 2026Neural Grammatical Error Correction for Romaniancited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.