# PROBE: Manipulation-Grounded Visual Question Answering with VLM Agents

arXiv: https://arxiv.org/abs/2608.17129 · HTML: https://mlsanity.com/paper/2608.17129/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 30 of its 33 references that are in the map.
- First posted on arXiv 17 Aug 2026.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)
- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 30 of 30)

- [Do As I Can, Not As I Say: Grounding Language in Robotic Affordances](https://mlsanity.com/paper/2204.01691/) (2022)
- [LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learning](https://mlsanity.com/paper/2306.03310/) (2023)
- [SAM 3: Segment Anything with Concepts](https://mlsanity.com/paper/2511.16719/) (2025)
- [AI2-THOR: An Interactive 3D Environment for Visual AI](https://mlsanity.com/paper/1712.05474/) (2017)
- [SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilities](https://mlsanity.com/paper/2401.12168/) (2024)
- [ProgPrompt: Generating Situated Robot Task Plans using Large Language Models](https://mlsanity.com/paper/2209.11302/) (2022)
- [ViperGPT: Visual Inference via Python Execution for Reasoning](https://mlsanity.com/paper/2303.08128/) (2023)
- [CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Tasks](https://mlsanity.com/paper/2112.03227/) (2021)
- [Habitat 2.0: Training Home Assistants to Rearrange their Habitat](https://mlsanity.com/paper/2106.14405/) (2021)
- [Ferret: Refer and Ground Anything Anywhere at Any Granularity](https://mlsanity.com/paper/2310.07704/) (2023)
- [RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robots](https://mlsanity.com/paper/2406.02523/) (2024)
- [Embodied Question Answering](https://mlsanity.com/paper/1711.11543/) (2017)
- [RoboPoint: A Vision-Language Model for Spatial Affordance Prediction for Robotics](https://mlsanity.com/paper/2406.10721/) (2024)
- [ManiSkill3: GPU Parallelized Robotics Simulation and Rendering for Generalizable Embodied AI](https://mlsanity.com/paper/2410.00425/) (2024)
- [MegaPose: 6D Pose Estimation of Novel Objects via Render & Compare](https://mlsanity.com/paper/2212.06870/) (2022)
- [ManiSkill: Generalizable Manipulation Skill Benchmark with Large-Scale Demonstrations](https://mlsanity.com/paper/2107.14483/) (2021)
- [VLABench: A Large-Scale Benchmark for Language-Conditioned Robotics Manipulation with Long-Horizon Reasoning Tasks](https://mlsanity.com/paper/2412.18194/) (2024)
- [RoboRefer: Towards Spatial Referring with Reasoning in Vision-Language Models for Robotics](https://mlsanity.com/paper/2506.04308/) (2025)
- [RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins](https://mlsanity.com/paper/2504.13059/) (2025)
- [CaP-X: A Framework for Benchmarking and Improving Coding Agents for Robot Manipulation](https://mlsanity.com/paper/2603.22435/) (2026)
- [From Pixels to Graphs: Open-Vocabulary Scene Graph Generation with Vision-Language Models](https://mlsanity.com/paper/2404.00906/) (2024)
- [GraspGen: A Diffusion-based Framework for 6-DOF Grasping with On-Generator Training](https://mlsanity.com/paper/2507.13097/) (2025)
- [Digital Twin Catalog: A Large-Scale Photorealistic 3D Object Digital Twin Dataset](https://mlsanity.com/paper/2504.08541/) (2025)
- [ESI-Bench: Towards Embodied Spatial Intelligence that Closes the Perception-Action Loop](https://mlsanity.com/paper/2605.18746/) (2026)
- [RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies](https://mlsanity.com/paper/2604.09860/) (2026)
- [SpaceTools: Tool-Augmented Spatial Reasoning via Double Interactive RL](https://mlsanity.com/paper/2512.04069/) (2025)
- [BOP Challenge 2024 on Model-Based and Model-Free 6D Object Pose Estimation](https://mlsanity.com/paper/2504.02812/) (2025)
- [MQA: Answering the Question via Robotic Manipulation](https://mlsanity.com/paper/2003.04641/) (2020)
- [Embodied Referring Expression for Manipulation Question Answering in Interactive Environment](https://mlsanity.com/paper/2210.02709/) (2022)
- [BOP-ASK: Object-Interaction Reasoning for Vision-Language Models](https://mlsanity.com/paper/2511.16857/) (2025)

## Built on by (most cited first, 1 of 1)

- [In-Context Learning for Robots: Methods and Applications](https://mlsanity.com/paper/2609.36012/) (2026)

## Abstract (arXiv)

Vision-language models (VLMs) have shown promising spatial reasoning capabilities from static visual inputs, where the evidence needed to answer a question is available in the provided views. However, in cluttered environments, answer-relevant evidence may be occluded rather than absent: an object may lie beneath a pile, be covered by another object, or have identifying information facing away from the camera. Answering such questions requires physical interaction to reveal the hidden evidence. We formalize this setting as Manipulation-Grounded Visual Question Answering (MG-VQA), where an agent answers a question about an initially cluttered scene by using manipulation as an intermediate evidence-gathering operation. We introduce MG-VQA-Bench, comprising 600 human-verified questions across four spatial reasoning tasks, and evaluate it in a cluttered tabletop interactive environment. Eight VLMs are evaluated under three levels of environment access: Direct (image only), Perception (pointing, segmentation, and scene graphs), and Manipulation (perception, grasping, and pushing). Across all models, Direct and Perception achieve average success rates of 36.2% and 37.6%, respectively, only slightly above an image-blind chance baseline of 32.7%. In contrast, Manipulation raises average success to 56.8% (40.8-83.0% across models). Stronger tool-calling VLMs, GPT~6~Astra (83.0%) and Gemini~3.8~Flash (69.3%), search persistently and re-ground after unsuccessful interactions, while weaker models often answer without gathering sufficient evidence or stop prematurely after failed actions. Our results highlight the need for VLM agents that use manipulation for persistent, physically grounded evidence gathering and recovery. Code and benchmark is available here: https://github.com/vineet2104/MG-VQA