Chain-of-Thought Degrades Visual Spatial Reasoning Capabilities of Multimodal LLMs
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915942165905408 |
|---|---|
| author | Kancheti, Sai Srinivas Kanade, Aditya Sanjiv Balasubramanian, Vineeth N. Ganu, Tanuja |
| author_facet | Kancheti, Sai Srinivas Kanade, Aditya Sanjiv Balasubramanian, Vineeth N. Ganu, Tanuja |
| contents | Multimodal Reasoning Models (MRMs) leveraging Chain-of-Thought (CoT) based thinking have revolutionized mathematical and logical problem-solving. However, we show that this paradigm struggles with generalized spatial intelligence. We perform a comprehensive evaluation of seventeen models across thirteen spatial benchmarks and identify a critical gap: CoT prompting consistently degrades performance in visual spatial reasoning. Furthermore, through a novel No-Image++ ablation, we demonstrate that MRMs and CoT prompted MLMs suffer from severe shortcut learning, and hallucinate visual details from textual priors even when the image is absent. These findings challenge the efficacy of text-only CoT for spatial tasks and underscore the need for vision-centric reasoning paradigms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_16060 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Chain-of-Thought Degrades Visual Spatial Reasoning Capabilities of Multimodal LLMs Kancheti, Sai Srinivas Kanade, Aditya Sanjiv Balasubramanian, Vineeth N. Ganu, Tanuja Computer Vision and Pattern Recognition Artificial Intelligence Multimodal Reasoning Models (MRMs) leveraging Chain-of-Thought (CoT) based thinking have revolutionized mathematical and logical problem-solving. However, we show that this paradigm struggles with generalized spatial intelligence. We perform a comprehensive evaluation of seventeen models across thirteen spatial benchmarks and identify a critical gap: CoT prompting consistently degrades performance in visual spatial reasoning. Furthermore, through a novel No-Image++ ablation, we demonstrate that MRMs and CoT prompted MLMs suffer from severe shortcut learning, and hallucinate visual details from textual priors even when the image is absent. These findings challenge the efficacy of text-only CoT for spatial tasks and underscore the need for vision-centric reasoning paradigms. |
| title | Chain-of-Thought Degrades Visual Spatial Reasoning Capabilities of Multimodal LLMs |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2604.16060 |