# PolyBridgeBench: Benchmarking Multimodal LLMs for Physics-Grounded Bridge Design

arXiv: https://arxiv.org/abs/2609.21493 · HTML: https://mlsanity.com/paper/2609.21493/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Chart understanding and anomaly detection, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 17 of its 23 references that are in the map.
- First posted on arXiv 18 Sep 2026.

## Placed in

- Multimodal LLMs / vision-language models / [Chart understanding and anomaly detection](https://mlsanity.com/mllm/chart-understanding-and-anomaly-detection/)

## Builds on (most cited first, 17 of 17)

- [PIQA: Reasoning about Physical Commonsense in Natural Language](https://mlsanity.com/paper/1911.11641/) (2019)
- [CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning](https://mlsanity.com/paper/1612.06890/) (2016)
- [SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilities](https://mlsanity.com/paper/2401.12168/) (2024)
- [Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces](https://mlsanity.com/paper/2412.14171/) (2024)
- [A Corpus for Reasoning About Natural Language Grounded in Photographs](https://mlsanity.com/paper/1811.00491/) (2018)
- [PlanBench: An Extensible Benchmark for Evaluating Large Language Models on Planning and Reasoning about Change](https://mlsanity.com/paper/2206.10498/) (2022)
- [PhysBench: Benchmarking and Enhancing Vision-Language Models for Physical World Understanding](https://mlsanity.com/paper/2501.16411/) (2025)
- [Physion: Evaluating Physical Prediction from Vision in Humans and Machines](https://mlsanity.com/paper/2106.08261/) (2021)
- [Embodied Agent Interface: Benchmarking LLMs for Embodied Decision Making](https://mlsanity.com/paper/2410.07166/) (2024)
- [PHYRE: A New Benchmark for Physical Reasoning](https://mlsanity.com/paper/1908.05656/) (2019)
- [RoboSpatial: Teaching Spatial Understanding to 2D and 3D Vision-Language Models for Robotics](https://mlsanity.com/paper/2411.16537/) (2024)
- [StepGame: A New Benchmark for Robust Multi-Hop Spatial Reasoning in Texts](https://mlsanity.com/paper/2204.08292/) (2022)
- [LEGO-Puzzles: How Good Are MLLMs at Multi-Step Spatial Reasoning?](https://mlsanity.com/paper/2503.19990/) (2025)
- [CAD-Assistant: Tool-Augmented VLLMs as Generic CAD Task Solvers](https://mlsanity.com/paper/2412.13810/) (2024)
- [Text-to-CAD Evaluation with CADTests](https://mlsanity.com/paper/2605.07807/) (2026)
- [BuildArena: A Physics-Aligned Interactive Benchmark of LLMs for Engineering Construction](https://mlsanity.com/paper/2510.16559/) (2025)
- [ORIGAMISPACE: Benchmarking Multimodal LLMs in Multi-Step Spatial Reasoning with Mathematical Constraints](https://mlsanity.com/paper/2511.18450/) (2025)

## Abstract (arXiv)

Multimodal large language models, or MLLMs, perform well at visual understanding and structured generation, yet these capabilities do not establish whether an engineering design will work when executed. Existing benchmarks assess spatial reasoning, structural validity, or physics-grounded construction, but they do not determine whether MLLMs can synthesize complete load-bearing structures and repair them after simulator execution exposes a failure. We introduce PolyBridgeBench, an executable benchmark for multimodal bridge design. A model receives a visual scene and structured engineering constraints and generates a complete node--member--material topology. Deterministic legality checks gate execution in a native dynamic physics simulation. Following an execution failure, the benchmark returns temporal visual evidence from the failed rollout and evaluates repair under a fixed interaction budget. Separate measurements of deterministic validity, dynamic functional success, and post-failure recovery identify the stage at which design fails. Experiments with six representative MLLMs across 189 levels expose a substantial gap between deterministic validity and dynamic success, pronounced sensitivity to material budgets, and limited post-failure recovery under the primary strict-budget setting.