# BilliardPhys-Bench: Benchmarking Physical Reasoning and Visual Dynamics of Multimodal LLMs

arXiv: https://arxiv.org/abs/2605.30900 · HTML: https://mlsanity.com/paper/2605.30900/

## 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 34 of its 40 references that are in the map.
- First posted on arXiv 29 May 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/)

## Builds on (most cited first, 30 of 34)

- [MiniCPM-V: A GPT-4V Level MLLM on Your Phone](https://mlsanity.com/paper/2408.01800/) (2024)
- [Cosmos World Foundation Model Platform for Physical AI](https://mlsanity.com/paper/2501.03575/) (2025)
- [mPLUG-Owl3: Towards Long Image-Sequence Understanding in Multi-Modal Large Language Models](https://mlsanity.com/paper/2408.04840/) (2024)
- [PHYRE: A New Benchmark for Physical Reasoning](https://mlsanity.com/paper/1908.05656/) (2019)
- [UGPhysics: A Comprehensive Benchmark for Undergraduate Physics Reasoning with Large Language Models](https://mlsanity.com/paper/2502.00334/) (2025)
- [Kinetix: Investigating the Training of General Agents through Open-Ended Physics-Based Control Tasks](https://mlsanity.com/paper/2410.23208/) (2024)
- [PhyX: Does Your Model Have the "Wits" for Physical Reasoning?](https://mlsanity.com/paper/2505.15929/) (2025)
- [WholeBodyVLA: Towards Unified Latent VLA for Whole-Body Loco-Manipulation Control](https://mlsanity.com/paper/2512.11047/) (2025)
- [SeePhys: Does Seeing Help Thinking? -- Benchmarking Vision-Based Physics Reasoning](https://mlsanity.com/paper/2505.19099/) (2025)
- [LLMPhy: Parameter-Identifiable Physical Reasoning Combining Large Language Models and Physics Engines](https://mlsanity.com/paper/2411.08027/) (2024)
- [Probing Scientific General Intelligence of LLMs with Scientist-Aligned Workflows](https://mlsanity.com/paper/2512.16969/) (2025)
- [Being-0: A Humanoid Robotic Agent with Vision-Language Models and Modular Skills](https://mlsanity.com/paper/2503.12533/) (2025)
- [Probing the Critical Point (CritPt) of AI Reasoning: a Frontier Physics Research Benchmark](https://mlsanity.com/paper/2509.26574/) (2025)
- [QuantiPhy: A Quantitative Benchmark Evaluating Physical Reasoning Abilities of Vision-Language Models](https://mlsanity.com/paper/2512.19526/) (2025)
- [ABench-Physics: Benchmarking Physical Reasoning in LLMs via High-Difficulty and Dynamic Physics Problems](https://mlsanity.com/paper/2507.04766/) (2025)
- [TSRBench: A Comprehensive Multi-task Multi-modal Time Series Reasoning Benchmark for Generalist Models](https://mlsanity.com/paper/2601.18744/) (2026)
- [PhysUniBench: A Multi-Modal Physics Reasoning Benchmark at Undergraduate Level](https://mlsanity.com/paper/2506.17667/) (2025)
- [DeepPHY: Benchmarking Agentic VLMs on Physical Reasoning](https://mlsanity.com/paper/2508.05405/) (2025)
- [WorldBench: Benchmarking Physical Understanding of World Models by Isolating Physics Concepts](https://mlsanity.com/paper/2601.21282/) (2026)
- [Real Time Trajectory Prediction Using Deep Conditional Generative Models](https://mlsanity.com/paper/1909.03895/) (2019)
- [ReplicationBench: Can AI Agents Replicate Astrophysics Research Papers?](https://mlsanity.com/paper/2510.24591/) (2025)
- [PhysicsMind: Sim and Real Mechanics Benchmarking for Physical Reasoning and Prediction in Foundational VLMs and World Models](https://mlsanity.com/paper/2601.16007/) (2026)
- [UniDriveDreamer: A Single-Stage Multimodal World Model for Autonomous Driving](https://mlsanity.com/paper/2602.02002/) (2026)
- [RSA-Bench: Benchmarking Audio Large Models in Real-World Acoustic Scenarios](https://mlsanity.com/paper/2601.10384/) (2026)
- [Black-Box vs. Gray-Box: A Case Study on Learning Table Tennis Ball Trajectory Prediction with Spin and Impacts](https://mlsanity.com/paper/2305.15189/) (2023)
- [SmallWorlds: Assessing Dynamics Understanding of World Models in Isolated Environments](https://mlsanity.com/paper/2511.23465/) (2025)
- [Digital Twin AI: Opportunities and Challenges from Large Language Models to World Models](https://mlsanity.com/paper/2601.01321/) (2026)
- [Egocentric Event-Based Vision for Ping Pong Ball Trajectory Prediction](https://mlsanity.com/paper/2506.07860/) (2025)
- [Embracing Evolution: A Call for Body-Control Co-Design in Embodied Humanoid Robot](https://mlsanity.com/paper/2510.03081/) (2025)
- [Uplifting Table Tennis: A Robust, Real-World Application for 3D Trajectory and Spin Estimation](https://mlsanity.com/paper/2511.20250/) (2025)

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

- [Rule-Constrained Assignment for Cue-Ball Identification in Broadcast Snooker](https://mlsanity.com/paper/2609.23450/) (2026)

## Abstract (arXiv)

Current multimodal models handle static image recognition well, but intuitive physical reasoning remains a weakness. Predicting how objects will move and interact from a single image is still difficult for these systems. We present BilliardPhys-Bench, a benchmark for physical reasoning in synthetic billiards environments. Its procedural engine generates randomized scenarios with friction and elastic collisions. The benchmark tests three abilities: (1) predicting ball-to-ball collisions, (2) reasoning about wall bounces, and (3) estimating final ball positions after motion stops. We evaluate recent MLLMs from the GPT, Claude, Gemini, and Qwen families. Performance drops as simulation time increases and scene geometry grows more complex. We also observe a consistent failure mode we call "stasis bias": when the correct physical outcome is harder to infer, models tend to predict no interaction. These findings show where current MLLMs break down on visual dynamics and point toward the need for better physical inductive biases in multimodal architectures.