# Tri-Bench: Stress-Testing VLM Reliability on Spatial Reasoning under Camera Tilt and Object Interference

arXiv: https://arxiv.org/abs/2512.08860 · HTML: https://mlsanity.com/paper/2512.08860/

## 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 4 of its 12 references that are in the map.
- First posted on arXiv 9 Dec 2025.

## 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, 4 of 4)

- [OmniSpatial: Towards Comprehensive Spatial Reasoning Benchmark for Vision Language Models](https://mlsanity.com/paper/2506.03135/) (2025)
- [Mind the Gap: Benchmarking Spatial Reasoning in Vision-Language Models](https://mlsanity.com/paper/2503.19707/) (2025)
- [AutoGPS: Automated Geometry Problem Solving via Multimodal Formalization and Deductive Reasoning](https://mlsanity.com/paper/2505.23381/) (2025)
- [VisioMath: Benchmarking Figure-based Mathematical Reasoning in LMMs](https://mlsanity.com/paper/2506.06727/) (2025)

## Abstract (arXiv)

Verifiable geometric reasoning is a critical component for trustworthy and controllable agentic AI. Despite impressive capabilities, Vision-Language Models (VLMs) often fail under realistic scene changes. We present Tri-Bench, a compact benchmark of planar triangle problems that isolates relative geometric reasoning while stressing two deployment-critical factors: camera pose (planar vs. tilted) and scene context via object interference (10 everyday objects). To test verifiability and control, we evaluate four recent VLMs using a single, fixed prompt whose guardrail explicitly describes a surrounding square border, enabling correct answers via homography. We evaluate six simple tasks over binary and continuous targets, and observe that the overall accuracy with respect to 3D ground truth is modest, ~69% on average (best ~75%, worst ~64%). The same responses align even more closely with 2D projections in the image plane, where mean accuracy is ~72%. All four VLMs consistently fail, with accuracy falling to ~0%, on recognizing minority shape classes (equilateral, isosceles, right-angled triangles). Additionally, overall VLM accuracy degrades by ~4.1% under camera tilt. This demonstrates that models fail to correctly utilize the explicit frame-of-reference hint provided in the prompt and default to 2D image plane cues. Finally, we find that object interference has no significant effect on VLM accuracy.