從傷害到助力：將情境推理示範轉化為推理語言模型的資產

摘要

近期，基于推理的大型语言模型（RLMs），尤其是那些通过验证器强化学习训练的模型，在少样本思维链（CoT）下的表现往往不如直接回答。我们利用DeepSeek-R1提供的高质量推理轨迹作为示例，重新审视了这一悖论，发现即使示例是最优的，增加示例数量仍会持续降低准确性。深入分析揭示了导致这一下降的两大机制：(i) 语义误导，即高文本相似性使模型将目标问题视为与示例相同，并逐字复制中间步骤；(ii) 策略转移失败，即模型难以提取有用的推理策略并将其应用于目标问题。基于这些发现，我们引入了“洞察至解决”（I2S），一种顺序测试时程序，它将示例转化为明确、可重用的洞察，并生成针对特定目标的推理轨迹；此外，推理过程可自我优化以确保连贯性和正确性（I2S+）。在多样化的基准测试上的广泛实验表明，I2S和I2S+在开源和闭源模型中均一致优于直接回答和测试时扩展基线。即便是对于GPT模型，我们的方法也大有裨益：在AIME'25上，GPT-4.1提升了+14.0%，o1-mini在AIME上提升了+2.7%，在GPQA上提升了+1.7%，这表明通过“洞察-优化-解决”框架，可以有效地利用上下文示例。

English

Recent reasoning LLMs (RLMs), especially those trained with verifier-based reinforcement learning, often perform worse with few-shot CoT than with direct answering. We revisit this paradox using high-quality reasoning traces from DeepSeek-R1 as demonstrations and find that adding more exemplars consistently degrades accuracy, even when demonstrations are optimal. A detailed analysis reveals two mechanisms behind this decline: (i) semantic misguidance, where high textual similarity leads the model to treat the target as the same as the exemplar and to copy intermediate steps verbatim; and (ii) strategy transfer failure, where the model struggles to extract useful reasoning strategies and apply them to target questions. Guided by these, we introduce Insight-to-Solve (I2S), a sequential test-time procedure that turns demonstrations into explicit, reusable insights and derives a target-specific reasoning trace; optionally, the reasoning is self-refined for coherence and correctness (I2S+). Extensive experiments on diverse benchmarks show that I2S and I2S+ consistently outperform both direct answering and test-time scaling baselines across open- and closed-source models. Even for GPT models, our method helps: on AIME'25, GPT-4.1 rises by +14.0%, and o1-mini improves by +2.7% on AIME and +1.7% on GPQA, indicating that in-context demonstrations can be harnessed effectively via insight-refine-solve framework.

從傷害到助力：將情境推理示範轉化為推理語言模型的資產

From Harm to Help: Turning Reasoning In-Context Demos into Assets for Reasoning LMs

摘要

Support