Tag
RoTA-Bench 是一个评测视觉-语言模型将 2D 观察转化为精确 3D 旋转控制能力的基准，通过交互式投影对齐任务揭示了显著的人机差距（GPT-5.6-Sol 仅 17% 成功率，而人类参与者 100% 完成）。
Imagine3D-LLM proposes teaching multimodal LLMs to imagine a compact 3D representation of a scene (decoded from learnable summary tokens as 3D Gaussian Splatting via photometric reconstruction loss) before answering multi-view spatial reasoning questions, outperforming prior 3D-aware MLLM approaches on multiple benchmarks.
Skill-3D is a framework that enables AI agents to learn scene-aware skills through self-evolving memory and skill libraries, significantly improving tool utilization in 3D spatial reasoning tasks (e.g., from 39% to 78% on VSI-Bench).
This paper proposes GASP, a framework that injects geometric priors into vision-language models via deep supervision with contrastive and depth consistency losses, achieving significant improvements on 3D spatial reasoning benchmarks without using 3D VQA data.
MetaSpatial is a reinforcement learning framework that enhances 3D spatial reasoning in vision-language models, enabling coherent and physically plausible 3D scene generation without hard-coded optimizations.