GeoThinker-$F$: Endogenous Predictive World Representations for Fast Spatial Intelligence

Paper recorded by Signals 4 on 2026-10-14 in cs.CV. Abstract reproduced from arXiv; link to the original below.

Published 2026-10-14 on arXiv · recorded by Signals 4 on 2026-10-15

Category: cs.CV · 计算机视觉 · first seen 2026-10-15

Abstract

Existing geometry-aware Multimodal Large Language Models (MLLMs) increasingly rely on external 3D encoders to provide structural priors, but executing these additional encoder paths at inference introduces substantial latency and computational overhead. We propose GeoThinker-$F$, an efficient framework that internalizes spatiotemporal perception by transforming causal MLLM hidden states into predi

Read on arXiv →