Research map / Large language models / Language, safety and interpretability
MaskGEC: Improving Neural Grammatical Error Correction via Dynamic Masking
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add dynamic mixed-substitution masking to source side (ablation) | Char-Transformer · NLPCC-2018 Task 2 | F0.5: 27.86 -> 36.97 | Helped |
| use dynamic masking instead of static masking (ablation) | MaskGEC · NLPCC-2018 Task 2 | F0.5: 36.35 -> 36.97 | Helped |
| add padding substitution only (ablation) | Char-Transformer · NLPCC-2018 Task 2 | F0.5: 27.86 -> 33.09 | Helped |
| add random substitution only (ablation) | Char-Transformer · NLPCC-2018 Task 2 | F0.5: 27.86 -> 32.37 | Helped |
| add word frequency substitution only (ablation) | Char-Transformer · NLPCC-2018 Task 2 | F0.5: 27.86 -> 30.11 | Helped |
| add homophone substitution only (ablation) | Char-Transformer · NLPCC-2018 Task 2 | F0.5: 27.86 -> 29.75 | Helped |
| mix all noising schemes rather than single scheme (ablation) | MaskGEC · NLPCC-2018 Task 2 | F0.5: 33.09 -> 36.97 | Helped |
| dynamic masking vs source-word dropout | Char-Transformer · NLPCC-2018 Task 2 | F0.5: 32.15 -> 36.97 | Helped |
| use BPE subword units instead of characters (ablation) | Transformer · NLPCC-2018 Task 2 | F0.5: 27.86 -> 24.94 | Hurt |
| vary substitution probability delta (best 0.3) (ablation) | MaskGEC · NLPCC-2018 Task 2 | F0.5 | Mixed |
Closest papers by meaning
- ALRMR-GEC: Adjusting Learning Rate Based on Memory Rate to Optimize the Edit Scorer for Grammatical Error Correction (AAAI 2025 · similarity 0.88)
- ScholarGEC: Enhancing Controllability of Large Language Model for Chinese Academic Grammatical Error Correction (AAAI 2025 · similarity 0.87)
- DSGram: Dynamic Weighting Sub-Metrics for Grammatical Error Correction in the Era of Large Language Models (AAAI 2025 · similarity 0.86)
- EXCGEC: A Benchmark for Edit-Wise Explainable Chinese Grammatical Error Correction (AAAI 2025 · similarity 0.85)
- Enhancing Large Language Model Performance with Gradient-Based Parameter Selection (AAAI 2025 · similarity 0.85)
- Learning from Mistakes: Self-correct Adversarial Training for Chinese Unnatural Text Correction (AAAI 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.