TY  - RPRT
TI  - GAP-MLLM: Geometry-Aligned Pre-training for Activating 3D Spatial Perception in Multimodal Large Language Models
AU  - Jiaxin Zhang
AU  - Junjun Jiang
AU  - Haijie Li
AU  - Youyu Chen
AU  - Kui Jiang
AU  - Dave Zhenyu Chen
PY  - 2026
UR  - https://arxiv.org/abs/2603.16461
ID  - 2603.16461
ER  -