无思考校准的思考：推理大型语言模型中的一种新上下文学习范式
Thinking with Nothinking Calibration: A New In-Context Learning Paradigm in Reasoning Large Language Models
AI总结：
提出一种名为JointThinking的新ICL范式，通过并行生成思考与无思考答案并在不一致时触发二次思考，显著提升RLLMs的推理性能与泛化能力。
摘要：
推理大型语言模型（RLLMs）近期通过结构化和多步推理展现了卓越的能力。尽管先前的研究主要集中在改进其训练和推理策略上，但其上下文学习（ICL）的潜力在很大程度上仍未被探索。为了填补这一空白，我们提出了无思考校准的思考（JointThinking），这是一种新的ICL范式，提示模型并行生成两个答案：一个处于思考模式，另一个处于无思考模式。仅当两个初始响应不一致时，才使用带有两个不同答案的单个提示触发第二轮思考。在多个推理基准上的广泛实验表明，JointThinking显著优于少样本思维链（CoT）、思考两次和多数投票。此外，它在分布内性能上实现了与基于训练的SOTA推理方法相当的表现，同时在分布外任务上大幅超越。我们进一步对校准机制进行了系统分析，展示了结构化思考多样性的重要性和一致性检查的益处。此外，我们观察到在第二次思考中，实际与理想推理之间的性能差距随着模型规模的增加而缩小，表明我们方法的强大可扩展性。最后，我们讨论了当前的局限性，并概述了RLLMs中未来ICL研究的有希望的方向。
英文摘要：
Reasoning large language models (RLLMs) have recently demonstrated remarkable capabilities through structured and multi-step reasoning. While prior research has primarily focused on improving their training and inference strategies, their potential for in-context learning (ICL) remains largely underexplored. To fill this gap, we propose Thinking with Nothinking Calibration (JointThinking), a new ICL paradigm that prompts the model to generate two answers in parallel: one in Thinking mode and the other in Nothinking mode. A second round of Thinking is triggered only when the two initial responses are inconsistent, using a single prompt with two different answers. Extensive experiments across multiple reasoning benchmarks demonstrate that JointThinking significantly outperforms few-shot chain-of-thought (CoT), thinking twice and majority voting. Moreover, it achieves comparable in-distribution performance to training-based SOTA reasoning method, while substantially outperforming on out-of-distribution tasks. We further conduct a systematic analysis of the calibration mechanism, showing the importance of structural thinking diversity and the benefits of consistency check. Additionally, we observe that the performance gap between actual and ideal reasoning narrows as model size increases in the second thinking, indicating the strong scalability of our approach. Finally, we discuss current limitations and outline promising directions for future ICL research in RLLMs.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
