arXiv cs.AI· Jiaru Qian, Yihong Dong, Xue Jiang, Chun-Wah Hsu, Zhi Jin, Ge Li·· 13 小时前AI 评分29
Diverse Parallel Thinking（DPT）：用路径多样性改进 LLM 并行思考
Parallel Thinking for Large Language Model with Path Diversity
AI 导读
研究者提出 Diverse Parallel Thinking（DPT），一种通过惩罚并行思考路径的前缀相似度来鼓励路径多样性的强化学习方法，缓解 RL 并行思考中推理路径的模式坍缩问题。在多个推理基准上，DPT 优于此前的并行思考方法，达到 SOTA，并验证路径多样性是提升并行思考的关键要素。代码已开源。
正文 · 原文
Abstract:Parallel thinking has emerged as a promising paradigm for enhancing the reasoning capabilities of Large Language Models (LLMs) by exploring multiple reasoning paths simultaneously. However, we identify that parallel thinking via Reinforcement Learning (RL) typically suffers from the mode collapse of reasoning paths, where parallel thinking trajectories degenerate into similar ones. This homogeneity undermines the fundamental motivation of parallel thinking, that is, maintaining multiple hypotheses concurrently to prevent premature loss into a suboptimal solution. In this paper, we propose Diverse Parallel Thinking (DPT), a simple yet effective RL approach for parallel thinking designed to encourage path diversity. DPT rewards the model to develop distinct reasoning strategies by penalizing the prefix similarity of parallel thinking paths, thereby enhancing a more comprehensive exploration of the reasoning space. We conduct various experiments and analyses across multiple reasoning benchmarks to demonstrate the effectiveness of DPT. Results show that DPT successfully outperforms previous parallel thinking methods, achieving state-of-the-art performance over baselines. We validate path diversity as a key element to improve parallel thinking. Our code is available at this https URL.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.29501 [cs.AI] |
| (or arXiv:2610.29501v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29501 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jiaru Qian [view email]
[v1]
Wed, 21 Oct 2026 09:51:42 UTC (1,544 KB)
来源：arXiv cs.AI · arxiv.org