S3-CoT：自采样简洁推理实现高效链式推理LLM
S3-CoT: Self-Sampled Succinct Reasoning Enables Efficient Chain-of-Thought LLMs
AI总结：
S3-CoT通过自采样简洁推理实现高效链式推理LLM，解决冗余推理问题，提升模型性能与泛化能力。
摘要：
大型语言模型（LLM）配备链式推理（CoT）能够实现强大的性能，并提供了解LLM行为的窗口。然而，最近的证据表明，CoT能力的提升往往伴随着冗余的推理过程，这引发了关键问题：LLM能否获得类似人类系统1推理的快速思考模式？为探索此问题，我们的研究提出了一种基于激活引导的自采样框架，用于高效CoT学习。我们的方法可以从目标LLM自身诱导出风格一致且长度可变的推理轨迹，而无需任何教师指导，从而缓解了基于SFT方法的核心瓶颈——高质量监督数据的稀缺性。通过过滤数据并使用黄金答案，我们对高效CoT学习进行了SFT，实现了（i）类人双认知系统和（ii）渐进式压缩课程。此外，我们探索了一种自进化模式，在此模式中，SFT仅由长度可变变体的预测一致数据驱动，消除了对黄金答案的需求。在数学基准测试以及医学等跨领域泛化测试中，我们的方法在通用和R1型LLM上均实现了稳定的提升。我们的数据和模型检查点可在https://github.com/DYR1/S3-CoT上找到。
英文摘要：
Large language models (LLMs) equipped with chain-of-thought (CoT) achieve strong performance and offer a window into LLM behavior. However, recent evidence suggests that improvements in CoT capabilities often come with redundant reasoning processes, motivating a key question: Can LLMs acquire a fast-thinking mode analogous to human System 1 reasoning? To explore this, our study presents a self-sampling framework based on activation steering for efficient CoT learning. Our method can induce style-aligned and variable-length reasoning traces from target LLMs themselves without any teacher guidance, thereby alleviating a central bottleneck of SFT-based methods-the scarcity of high-quality supervision data. Using filtered data by gold answers, we perform SFT for efficient CoT learning with (i) a human-like dual-cognitive system, and (ii) a progressive compression curriculum. Furthermore, we explore a self-evolution regime in which SFT is driven solely by prediction-consistent data of variable-length variants, eliminating the need for gold answers. Extensive experiments on math benchmarks, together with cross-domain generalization tests in medicine, show that our method yields stable improvements for both general and R1-style LLMs. Our data and model checkpoints can be found at https://github.com/DYR1/S3-CoT.
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。
