视觉工具使用强化学习究竟在学习什么？解构工具诱导效应与内在效应以实现作物和缩放
What Does Vision Tool-Use Reinforcement Learning Really Learn? Disentangling Tool-Induced and Intrinsic Effects for Crop-and-Zoom
AI总结：
本文研究了视觉工具使用强化学习在作物和缩放任务中的学习机制，通过引入MED框架解耦内在能力变化与工具诱导效应，发现改进主要由内在学习驱动，而工具使用强化学习主要减少工具诱导的负面影响，而非掌握工具。
摘要：
视觉工具使用强化学习（RL）可以为视觉语言模型提供如作物和缩放等视觉操作，从而实现显著性能提升，但尚不清楚这些提升是源于工具使用能力的改进还是内在能力的演变。我们引入MED（测量-解释-诊断），一种由粗到细的框架，用于解耦内在能力变化与工具诱导效应，将工具诱导的性能差异分解为增益和损害项，并探测驱动其演变的机制。在作物和缩放设置中，对两个具有不同工具先验的VLMs和六个基准测试的检查点级分析显示，改进主要由内在学习驱动，而工具使用RL主要减少工具诱导的损害（例如更少的调用诱导错误和更弱的工具模式干扰），并在工具基于的内在失败修正方面取得有限进展。总体而言，在本文研究的作物和缩放设置中，当前的视觉工具使用RL学习的是安全地与工具共存，而非掌握工具。
英文摘要：
Vision tool-use reinforcement learning (RL) can equip vision language models with visual operators such as crop-and-zoom and achieves strong performance gains, yet it remains unclear whether these gains are driven by improvements in tool use or evolving intrinsic capabilities. We introduce MED (Measure--Explain--Diagnose), a coarse-to-fine framework that disentangles intrinsic capability changes from tool-induced effects, decomposes the tool-induced performance difference into gain and harm terms, and probes the mechanisms driving their evolution. Across checkpoint-level analyses in the crop-and-zoom setting on two VLMs with different tool priors and six benchmarks, we find that improvements are dominated by intrinsic learning, while tool-use RL mainly reduces tool-induced harm (e.g., fewer call-induced errors and weaker tool schema interference) and yields limited progress in tool-based correction of intrinsic failures. Overall, in the crop-and-zoom setting studied here, current vision tool-use RL learns to coexist safely with tools rather than master them.
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。
