SpatialClaw: Rethinking Action Interface for Agentic Spatial Reasoning

• URL: http://arxiv.org/abs/2606.13673v1
• Date: Thu, 11 Jun 2026 17:59:36 GMT
• Title: SpatialClaw: Rethinking Action Interface for Agentic Spatial Reasoning
• Abstract summary: We study how the design of this interface shapes the agent's capacity for open-ended spatial reasoning.<n>We propose SpatialClaw, a training-free framework for spatial reasoning that adopts code as the action interface.
• Score: 99.49739831030388
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Spatial reasoning, the ability to determine where objects are, how they relate, and how they move in 3D, remains a fundamental challenge for vision-language models (VLMs). Tool-augmented agents attempt to address this by augmenting VLMs with specialist perception modules, yet their effectiveness is bounded by the action interface through which those tools are invoked. In this work, we study how the design of this interface shapes the agent's capacity for open-ended spatial reasoning. Existing spatial agents either employ single-pass code execution, which commits to a full analysis strategy before any intermediate result is observed, or rely on a structured tool-call interface that often offers less flexibility for freely composing operations or tailoring the analysis to each task. Both designs offer limited flexibility for open-ended, complex 3D/4D spatial reasoning. We therefore propose SpatialClaw, a training-free framework for spatial reasoning that adopts code as the action interface. SpatialClaw maintains a stateful Python kernel pre-loaded with input frames and a suite of perception and geometry primitives, letting a VLM-backed agent write one executable cell per step conditioned on all prior outputs, enabling the agent to flexibly compose and manipulate perception results and adapt its analysis to both intermediate text and visual observations and the demands of each problem. Evaluated across 20 spatial reasoning benchmarks spanning a broad range of static and dynamic 3D/4D spatial reasoning tasks, SpatialClaw achieves 59.9% average accuracy, outperforming the recent spatial agent by +11.2 points, with consistent gains across six VLM backbones from two model families without any benchmark- or model-specific adaptation.

Related papers

• Chain of Spatial Thoughts: Modality-Agnostic Spatial Grounding for Vision Language Models [16.417990770250146]
  We introduce Space Tokens, a lightweight, architecture-agnostic framework that equips vision-language models with explicit continuous spatial representations.<n>By distilling scene-level 3D geometry and object-centric spatial attributes into continuous latent tokens, our method enables these modalities to be directly incorporated into a chain-of-thought reasoning process.<n>Results show that continuous spatial tokens provide an effective, interpretable, and computationally efficient mechanism for integrating geometric reasoning into large vision-language models.
  arXiv  Detail & Related papers  (2026-08-10T22:20:07Z)
• SpaceEra++: A Unified Framework Towards 3D Spatial Reasoning in Video [88.79625979053873]
  We propose a novel framework, SpaceEra, for pre-trained vision-language models (VLMs)<n>We extend the original framework into a comprehensive system, termed SpaceEra++, which spans data construction, model design, training optimization, and prompting inference.<n>In addition, to enhance spatial reasoning, we develop SpaceAlign, which enforces pairwise object constraints by jointly exploiting absolute coordinates and relative spatial relations.
  arXiv  Detail & Related papers  (2026-07-02T06:56:29Z)
• Agentic RAG-VLM: Affordance-Aware Retrieval-Augmented Generation with Self-Reflective Planning for Robotic Grasping [19.599594594020235]
  We present Agentic RAG-VLM, a unified framework that bridges VLM-based semantic understanding and physically grounded grasp execution.<n>RAG-VLM integrates retrieval-augmented generation (RAG) with vision-language models (VLMs) and agentic self-reflective planning.<n>RAG-VLM achieves 78.3 percent overall success, a 53.3 percentage-point absolute gain over VLM-only baselines.
  arXiv  Detail & Related papers  (2026-06-30T06:30:22Z)
• S-Agent: Spatial Tool-Use Elicits Reasoning for Spatial Intelligence [77.5121349100339]
  textscS-Agent reshapes spatial perception into scene-centric understanding beyond frame-centric recognition.<n>textscS-Agent casts the VLM as a semantic planner that decides what evidence is needed.<n>textscS-Agent consistently improves both open-source and closed-source VLMs in a training-free manner.
  arXiv  Detail & Related papers  (2026-06-18T17:34:55Z)
• SpatialSV: Internalizing Interpretable 3D Spatial Awareness in MLLMs via Task-Oriented Visual Supervision [12.336413790708738]
  spatial intelligence of large language models (MLLMs) is crucial for understanding and interacting with the 3D world.<n>We propose SpatialSV, a framework designed to internalize robust 3D spatial awareness within MLLMs while simultaneously offering inherent interpretability.
  arXiv  Detail & Related papers  (2026-06-18T08:09:32Z)
• GeoHAT: Geometry-Adaptive Hybrid Action Transformer for Mobile Manipulation [6.488530751190965]
  Whole-body mobile manipulation requires coordinating mobile base and manipulator under shifting viewpoints.<n>We present GeoHAT, an end-to-end diffusion-based framework built on a simple principle.<n>Experiments on the ManiSkill-HAB simulation benchmark demonstrate that GeoHAT achieves a 79.3% mean success rate.
  arXiv  Detail & Related papers  (2026-06-11T14:25:09Z)
• The Art of Interrogation: Consistency Amplifies Factuality in Spatial Reasoning [65.67266333751569]
  Current Large Reasoning Models (LRMs) exhibit remarkable general capabilities but significantly underperform in spatial reasoning tasks.<n>We propose a self-supervised reinforcement learning framework that targets the internal reasoning process without requiring ground-truth annotations.<n>We show that this label-free consistency training approaches the accuracy of models trained with ground-truth supervision and achieves similar generalization across diverse tasks and data domains.
  arXiv  Detail & Related papers  (2026-06-10T10:50:06Z)
• AffordanceVLA: A Vision-Language-Action Model Empowering Action Generation through Affordance-Aware Understanding [55.535374902204296]
  AffordanceVLA is a unified framework that introduces structured affordance forecasting as a task-oriented intermediate representation.<n>We show that AffordanceVLA achieves strong performance across diverse manipulation scenarios.
  arXiv  Detail & Related papers  (2026-06-04T13:28:51Z)
• AFUN: Towards an Affordance Foundation Model for Functionality Understanding [12.890216832485647]
  We present ourmodel, a step towards an affordance foundation model for functionality understanding.<n>We build a large-scale standardized data pipeline that converts heterogeneous robot, human, simulation, and real-world scan data into a shared affordance schema.<n>Ourmodel outperforms all baselines by a large margin across 8 test sets from 4 benchmarks.
  arXiv  Detail & Related papers  (2026-06-01T17:50:16Z)
• ViSRA: A Video-based Spatial Reasoning Agent for Multi-modal Large Language Models [38.91282173333918]
  ViSRA is a training-free framework to probe the spatial reasoning mechanism of MLLMs.<n>It elicits spatial reasoning in a modular and curated manner by leveraging explicit spatial information from expert models.<n>It offers two key advantages: (1) human-aligned and transferable 3D understanding rather than task-specific overfitting; and (2) no post-training computational cost along with heavy manual curation datasets.
  arXiv  Detail & Related papers  (2026-05-11T07:20:09Z)
• Towards Foundation Models for 3D Scene Understanding: Instance-Aware Self-Supervised Learning for Point Clouds [53.82500407523346]
  PointINS is an instance-oriented self-supervised framework that enriches point cloud representations through geometry-aware learning.<n>PointINS achieves on average +3.5% mAP improvement for indoor instance segmentation and +4.1% PQ gain for outdoor panoptic segmentation.
  arXiv  Detail & Related papers  (2026-03-26T08:31:06Z)
• Universal Pose Pretraining for Generalizable Vision-Language-Action Policies [83.39008378156647]
  Existing Vision-Language-Action (VLA) models often suffer from feature collapse and low training efficiency.<n>We propose Pose-VLA, a decoupled paradigm that separates VLA training into a pre-training phase for extracting universal 3D spatial priors.<n>Our framework follows a two-stage pre-training pipeline, establishing fundamental spatial grounding via poses followed by motion alignment.
  arXiv  Detail & Related papers  (2026-02-23T11:00:08Z)
• Generalized Decoupled Learning for Enhancing Open-Vocabulary Dense Perception [71.26728044621458]
  DeCLIP is a novel framework that enhances CLIP by decoupling the self-attention module to obtain content'' and context'' features respectively.<n>It consistently achieves state-of-the-art performance across a broad spectrum of tasks, including 2D detection and segmentation, 3D instance segmentation, video instance segmentation, and 6D object pose estimation.
  arXiv  Detail & Related papers  (2025-08-15T06:43:51Z)
• AnchorDP3: 3D Affordance Guided Sparse Diffusion Policy for Robotic Manipulation [8.603450327406879]
  AnchorDP3 is a diffusion policy framework for dual-arm robotic manipulation.<n>It is trained on large-scale, procedurally generated simulation data.<n>It achieves a 98.7% average success rate in the RoboTwin benchmark.
  arXiv  Detail & Related papers  (2025-06-24T03:03:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.