arXiv cs.CL· Taehoon Kim, Seunggeun Cho, Dongsu Han·· 10 天前AI 评分27
通过交叉反馈与连贯策展的协作推理蒸馏
Collaborative Reasoning Distillation via Cross-Feedback and Coherent Curation
AI 导读
研究提出协作推理蒸馏（CRD）框架，通过教师模型交叉反馈、逐步质量评估和连贯步骤拼接三项创新，将推理能力蒸馏到小模型。CRD-4B 在 MATH-500 上达 97.3%、AIME'25 上达 70.3%，仅用 50K 训练样本，数据集规模比同类模型小最多 12 倍。
正文 · 原文
Abstract:Reasoning capabilities are critical for advancing Large Language Models, yet current approaches either require massive computational budgets or struggle to effectively distill reasoning to smaller models. Standard distillation methods rely on outcome-based rewards, failing to distinguish between sound reasoning and lucky guesses. We propose Collaborative Reasoning Distillation (CRD), a framework that enhances reasoning in compact models through three innovations: (1) interactive cross-feedback where teachers iteratively critique each other's reasoning, (2) fine-grained step-wise quality assessment capturing logical validity independent of final answers, and (3) coherence-aware step stitching that synthesizes complementary strengths. Students are trained via Reasoning Quality Optimization (RQO) with budget constraints. Our model, CRD-4B, achieves 97.3% on MATH-500 and 70.3% on AIME'25, surpassing baselines while using only 50K training examples, up to 12 times smaller than the datasets of comparable models.
| Comments: | Accepted at NeurIPS 2026 |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.09587 [cs.LG] |
| (or arXiv:2610.09587v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.09587 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Taehoon Kim [view email]
[v1]
Wed, 7 Oct 2026 07:31:33 UTC (429 KB)
来源：arXiv cs.CL · arxiv.org