Preprint
Oct 2026
Spatial Latent Reasoning for Embodied Reference Understanding
Spatial Latent Reasoning (SLR) is proposed, a framework that structures this supervision around an ordered sequence of geometric and visual states supervised by fingertip position and pointing direction, followed by four states aligned with target-region features.
Ling Li, Jian-Hui Zhong, Wei Liu et al.
· 0 citations