# Can Vision-Language Models See Squares? Text-Recognition Mediates Spatial Reasoning Across Three Model Families

arXiv: https://arxiv.org/abs/2602.15950 · HTML: https://mlsanity.com/paper/2602.15950/

## 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 13 of its 13 references that are in the map.
- First posted on arXiv 17 Feb 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, 13 of 13)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMs](https://mlsanity.com/paper/2401.06209/) (2024)
- [SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilities](https://mlsanity.com/paper/2401.12168/) (2024)
- [SpatialRGPT: Grounded Spatial Reasoning in Vision Language Models](https://mlsanity.com/paper/2406.01584/) (2024)
- [Visual Spatial Reasoning](https://mlsanity.com/paper/2205.00363/) (2022)
- [Teaching CLIP to Count to Ten](https://mlsanity.com/paper/2302.12066/) (2023)
- [From CLIP to DINO: Visual Encoders Shout in Multi-modal Large Language Models](https://mlsanity.com/paper/2310.08825/) (2023)
- [Contrastive Localized Language-Image Pre-Training](https://mlsanity.com/paper/2410.02746/) (2024)
- [DocVLM: Make Your VLM an Efficient Reader](https://mlsanity.com/paper/2412.08746/) (2024)
- [iVISPAR -- An Interactive Visual-Spatial Reasoning Benchmark for VLMs](https://mlsanity.com/paper/2502.03214/) (2025)
- [Your Vision-Language Model Can't Even Count to 20: Exposing the Failures of VLMs in Compositional Counting](https://mlsanity.com/paper/2510.04401/) (2025)
- [Can CLIP Count Stars? An Empirical Study on Quantity Bias in CLIP](https://mlsanity.com/paper/2409.15035/) (2024)
- [VITRIX-CLIPIN: Enhancing Fine-Grained Visual Understanding in CLIP via Instruction Editing Data and Long Captions](https://mlsanity.com/paper/2508.02329/) (2025)

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

- [Grid2Matrix: Revealing Digital Agnosia in Vision-Language Models](https://mlsanity.com/paper/2604.09687/) (2026)

## Abstract (arXiv)

We present a simple experiment that exposes a fundamental limitation in vision-language models (VLMs): the inability to accurately localize filled cells in binary grids when those cells lack textual identity. We generate fifteen 15x15 grids with varying density (10.7%-41.8% filled cells) and render each as two image types -- text symbols (. and #) and filled squares without gridlines -- then ask three frontier VLMs (Claude Opus, ChatGPT 5.2, and Gemini 3 Thinking) to transcribe them. In the text-symbol condition, Claude and ChatGPT achieve approximately 91% cell accuracy and 84% F1, while Gemini achieves 84% accuracy and 63% F1. In the filled-squares condition, all three models collapse to 60-73% accuracy and 29-39% F1. Critically, all conditions pass through the same visual encoder -- the text symbols are images, not tokenized text. The text-vs-squares F1 gap ranges from 34 to 54 points across models, demonstrating that VLMs behave as if they possess a high-fidelity text-recognition pathway for spatial reasoning that dramatically outperforms their native visual pathway. Each model exhibits a distinct failure mode in the squares condition -- systematic under-counting (Claude), massive over-counting (ChatGPT), and template hallucination (Gemini) -- but all share the same underlying deficit: severely degraded spatial localization for non-textual visual elements.