模型越强越难听话，数学推理能力提升反而削弱指令遵循能力。
Scaling Reasoning, Losing Control: Evaluating Instruction Following in Large Reasoning Models
- 构建专门评测数学推理中指令遵循的基准MathIF。
- 模型越擅长推理，越容易忽略用户指令，尤其在长输出时更严重。
- 简单修复方法能恢复服从性，但会牺牲推理性能。
指令遵循是使大语言模型与用户意图对齐的关键。尽管近期面向推理的模型在复杂数学问题上表现优异，其对自然语言指令的遵守程度仍缺乏系统评估。本文提出MathIF，一个专用于数学推理任务的指令遵循评测基准。实证分析揭示：提升推理能力与保持可控性之间存在持续矛盾——推理越强的模型越难以遵守用户指令。特别是经过提炼的长链思维或推理导向强化学习训练的模型，在生成长度增加时指令遵循能力显著下降。此外，我们发现简单干预可部分恢复服从性，但代价是推理性能降低。这些发现揭示了当前大模型训练范式的根本矛盾，呼吁开发更具指令感知的推理模型。代码与数据已开源：https://github.com/TingchenFu/MathIF。
原文摘要 · Abstract (English)
Instruction-following is essential for aligning large language models (LLMs) with user intent. While recent reasoning-oriented models exhibit impressive performance on complex mathematical problems, their ability to adhere to natural language instructions remains underexplored. In this work, we introduce MathIF, a dedicated benchmark for evaluating instruction-following in mathematical reasoning tasks. Our empirical analysis reveals a consistent tension between scaling up reasoning capacity and maintaining controllability, as models that reason more effectively often struggle to comply with user directives. We find that models tuned on distilled long chains-of-thought or trained with reasoning-oriented reinforcement learning often degrade in instruction adherence, especially when generation length increases. Furthermore, we show that even simple interventions can partially recover obedience, though at the cost of reasoning performance. These findings highlight a fundamental tension in current LLM training paradigms and motivate the need for more instruction-aware reasoning models. We release the code and data at https://github.com/TingchenFu/MathIF.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。