# JRDB-AVR: An Active Visual Reasoning Benchmark for Embodied Agents in Real-World Environments

arXiv: https://arxiv.org/abs/2609.35032 · HTML: https://mlsanity.com/paper/2609.35032/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 23 of its 63 references that are in the map.
- First posted on arXiv 28 Sep 2026.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 23 of 23)

- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models](https://mlsanity.com/paper/2304.10592/) (2023)
- [HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face](https://mlsanity.com/paper/2303.17580/) (2023)
- [ViperGPT: Visual Inference via Python Execution for Reasoning](https://mlsanity.com/paper/2303.08128/) (2023)
- [ToRA: A Tool-Integrated Reasoning Agent for Mathematical Problem Solving](https://mlsanity.com/paper/2309.17452/) (2023)
- [VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos](https://mlsanity.com/paper/2405.19209/) (2024)
- [MindCube: Spatial Mental Modeling from Limited Views](https://mlsanity.com/paper/2506.21458/) (2025)
- [VTool-R1: VLMs Learn to Think with Images via Reinforcement Learning on Multimodal Tool Use](https://mlsanity.com/paper/2505.19255/) (2025)
- [DrVideo: Document Retrieval Based Long Video Understanding](https://mlsanity.com/paper/2406.12846/) (2024)
- [Explain Before You Answer: A Survey on Compositional Visual Reasoning](https://mlsanity.com/paper/2508.17298/) (2025)
- [Zoom-Refine: Boosting High-Resolution Multimodal Understanding via Localized Zoom and Self-Refinement](https://mlsanity.com/paper/2506.01663/) (2025)
- [Unleashing the Potential of Multimodal LLMs for Zero-Shot Spatio-Temporal Video Grounding](https://mlsanity.com/paper/2509.15178/) (2025)
- [JRDB-Traj: A Dataset and Benchmark for Trajectory Forecasting in Crowds](https://mlsanity.com/paper/2311.02736/) (2023)
- [An Empirical Study on How Video-LLMs Answer Video Questions](https://mlsanity.com/paper/2508.15360/) (2025)
- [VIEW2SPACE: Studying Multi-View Visual Reasoning from Sparse Observations](https://mlsanity.com/paper/2603.16506/) (2026)
- [Probabilistic Active Goal Recognition](https://mlsanity.com/paper/2507.21846/) (2025)
- [Memory-Anchored Multimodal Reasoning for Explainable Video Forensics](https://mlsanity.com/paper/2508.14581/) (2025)
- [JRDB-Pose3D: A Multi-person 3D Human Pose and Shape Estimation Dataset for Robotics](https://mlsanity.com/paper/2602.03064/) (2026)
- [Mini-BEHAVIOR-Gran: Revealing U-Shaped Effects of Instruction Granularity on Language-Guided Embodied Agents](https://mlsanity.com/paper/2604.17019/) (2026)
- [What We Talk About When We Talk About LLM Planning: Evidence for Two Distinct Planning Abilities](https://mlsanity.com/paper/2607.11197/) (2026)

## Abstract (arXiv)

In complex embodied visual reasoning scenarios, an agent often has only a limited field of view, and the evidence needed to answer a question may be distributed across time, viewpoint, and interacting objects. A model may therefore give a plausible answer without ever observing the relevant object, time, or view that supports it. Current visual reasoning benchmarks largely evaluate passive observations and final answers, overlooking settings that require active reasoning and evidence acquisition. We introduce JRDB-AVR, a benchmark derived from existing real-world JRDB robotics data through a structured question-generation engine that turns this gap into an explicit evaluation: an embodied agentic system receives a visual reasoning question, requests bounded observations by timestamp and viewing angle, and is evaluated on both the final answer and the grounded visual evidence supporting it. The benchmark contains diverse questions over multiple real-world environments involving temporal search, viewpoint selection, and human-oriented compositional reasoning. We also introduce JRDB-AVR-Agent, a reference active reasoning agentic method that maintains an explicit observation-grounded graph-based world model and answers through solving. Experiments reveal a substantial gap between answer accuracy and evidence accuracy in current baselines, showing that current VLMs can produce unsupported correct answers and that active evidence-aware evaluation is necessary for embodied visual reasoning. Code and benchmark are available at https://github.com/ControlNet/JRDB-AVR.