Research map / Large language models / Agents, code and math
Intuitive Thinking: Expanding Large Language Models’ Thinking for Rapid Decision-Making on Candidate Corrections in Chinese Grammar Error Correction
Lintao Long, Ruizhang Huang, Ruina Bai, Yongbin Qin, Qihang Fu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| inject LLM token confidence into correction embeddings of the lightweight ExIT model (ablation) | ExIT (Qwen2.5-14B) · FCGEC-test | F0.5: 50.65 -> 58.63 | Improved |
| add LLM alignment loss on top of confidence injection (ablation) | ExIT (Qwen2.5-14B) · FCGEC-test | F0.5: 58.63 -> 59.12 | Improved |
| add LLM alignment loss alone, without confidence injection (ablation) | ExIT (Qwen2.5-14B) · FCGEC-test | F0.5: 50.65 -> 50.67 | No clear effect |
| apply ExIT reranking over sampled candidates | Qwen2.5-14B fine-tuned LLM · FCGEC-test | F0.5: 52.98 -> 59.12 | Improved |
| set LLM sampling temperature to 0 (greedy) | Qwen2.5-14B fine-tuned LLM · FCGEC-test | F0.5: 52.98 -> 56.05 | Improved |
| ExIT instead of greedy decoding at temperature 0 | Qwen2.5-14B fine-tuned LLM · FCGEC-test | F0.5: 56.05 -> 59.12 | Improved |
| train 14B with ExIT on small data instead of 7B on large-scale data | LLM CGEC training · FCGEC-test | F0.5 | Improved |
| apply ExIT trained on Qwen2.5-14B to other LLMs | ExIT · FCGEC-test, other LLMs | F0.5 | Improved |
| use ExIT over Qwen2-7B-EPO on FCGEC-test | Qwen2-7B · FCGEC-test | F0.5: 59.63 (EPO) vs 58.50 (ExIT) | No clear effect |
| scale base LLM from 7B to 14B in ExIT | ExIT (Qwen2.5) · FCGEC-test | F0.5: 56.17 -> 59.12 | Improved |
Closest papers by meaning
- ScholarGEC: Enhancing Controllability of Large Language Model for Chinese Academic Grammatical Error Correction (AAAI 2025 · similarity 0.92)
- EXCGEC: A Benchmark for Edit-Wise Explainable Chinese Grammatical Error Correction (AAAI 2025 · similarity 0.89)
- Chinese Spelling Correction as Rephrasing Language Model (AAAI 2024 · similarity 0.86)
- DSGram: Dynamic Weighting Sub-Metrics for Grammatical Error Correction in the Era of Large Language Models (AAAI 2025 · similarity 0.85)
- Can LLMs Correct Themselves? A Benchmark of Self-Correction in LLMs (NeurIPS 2025 · similarity 0.85)
- Gaining Wisdom from Setbacks: Aligning Large Language Models via Mistake Analysis (ICLR 2024 · similarity 0.85)
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.
