SpatialCLI：先使用空间工具学习推理，再脱离工具进行推理

摘要

视觉语言模型（VLMs）越来越多地被用于具身智能体中，以解释视觉输入、推理空间关系，并基于该推理做出任务级决策。然而，一个根本性的能力错配依然存在：通用VLM能够推理整体任务，却往往遗漏决定成败的视觉细节；而专业视觉模型虽能捕捉这些细节，却无法将其转化为任务级决策。在本工作中，我们提出了SpatialCLI，一个教导VLM使用空间工具进行推理并逐步内化这些工具所提供的专业感知能力的框架。SpatialCLI分为三个阶段：（1）调用（Call）阶段将专业视觉模型作为空间工具提供给VLM，以增强其感知；（2）学习（Learn）阶段使用冷启动SFT和智能体强化学习来改进工具使用；（3）内化（Internalize）阶段将成功的工具使用轨迹言语化，从而内化专业感知能力。我们还引入了SpatialCLI-Bench，这是一个包含516个样本的基准，用于评估涵盖定位、分割、深度和位姿的组合感知能力。在MindCube上，SpatialCLI将Qwen3-VL-8B-Instruct从29.3%提升至84.6%（使用工具时），超过了使用工具的GPT-5.6 Sol（72.1%），同时在内化后无工具情况下仍保持73.8%的表现。

English

Vision-language models (VLMs) are increasingly used in embodied agents to interpret visual inputs, reason about spatial relationships, and make task-level decisions based on that reasoning. However, a fundamental capability mismatch remains: general VLMs can reason about the overall task but often miss the visual details that determine success, while specialist vision models can capture those details but cannot translate them into task-level decisions. In this work, we propose SpatialCLI, a framework that teaches VLMs to reason with spatial tools and progressively internalize the specialist perceptual capabilities they provide. SpatialCLI proceeds in three stages: (1) Call exposes specialist vision models as spatial tools to augment the VLM's perception; (2) Learn uses Cold-Start SFT and agentic RL to improve tool use; and (3) Internalize verbalizes successful tool-use trajectories to internalize specialist perceptual capabilities. We further introduce SpatialCLI-Bench, a 516-example benchmark for compositional perception across localization, segmentation, depth, and pose. On MindCube, SpatialCLI raises Qwen3-VL-8B-Instruct from 29.3% to 84.6% with tools, surpassing GPT-5.6 Sol with tools (72.1%), while retaining 73.8% without tools after internalization.