Thinking with Images without Coordinates

arXiv:2610.14265v1 Announce Type: new Abstract: Recent advances in multi-modal large language models have shifted toward the "thinking with images" paradigm to improve reasoning faithfulness, yet existing methods rely heavily on explicit coordinate- based grounding. We identify two fundamental limitations in this approach, i.e., a semantic gap where sparse regressions miss high-dimensional latent structures, and a resolution gap due to the granularity mismatch between patch-level tokens and pixel-level coordinates. To address these, we propose TiCO (Thinking with images without COordinates). It reformulates visual reasoning as a semantic evidence discovery process within the latent space, and evidence is injected as token-level regions without pixel coordinates. Specifically, TiCO utilizes a dedicated probe token to estimate similarity distributions and performs importance sampling over spatially connected components to generate diverse, semantically dominant evidence rollouts. Extensive experiments across multiple high-resolution and general visual reasoning benchmarks demonstrate that our TiCO consistently outperforms state-of-the-art coordinate-based methods with significantly improved inference efficiency.

Related

• DeepSketcher: Internalizing Visual Manipulation for Multimodal Reasoning
• Test-time Scaling over Perception: Resolving the Grounding Paradox in Thinking with Images
• SpatialForge: Bootstrapping 3D-Aware Spatial Reasoning from Open-World 2D Images

Source: arXiv cs.CV | 2026-10-14