5 papers
PolicyAlign: Direct Policy-Based Safety Alignment for Large Language Models
Chang Wu, Junfeng Fang, Houcheng Jiang +5
Safety alignment of large language models (LLMs) typically depends on high-quality supervision data, such as safe demonstrations or preference pairs. However, in real-world deploym…
DriveStack-VLA: Render-Teacher Alignment for BEV-Based DeepStack Vision-Language-Action Model
Jingke Wang, Zhenru Zhao, Shuangming Lei +8
Vision-Language-Action driving models convert a pretrained Vision-Language Model into a driving policy, allowing them to use world knowledge and follow language guidances. However,…
RegimeVGGT: Layer-Wise Spatially Preserving Redundancy Removal for Visual Geometry Grounded Transformer
Jinhao You, Shuo Lyu, Zhuohang Lyu +5
Visual Geometry Grounded Transformer (VGGT) recovers dense 3D scene structure from multi-view images in one forward pass, but quadratic cross-frame attention limits its scalability…
SafeDojo: Safe Reinforcement Learning for VLA via Interactive World Model
Kai Tang, Peidong Jia, Zhong Chu +15
Safe control is a prerequisite for real-world embodied intelligence, for which safe reinforcement learning has emerged as a promising paradigm. However, existing safe reinforcement…
TC-IDM: Grounding Video Generation for Executable Zero-shot Robot Motion
Weishi Mi, Yong Bao, Xiaowei Chi +7
The vision-language-action (VLA) paradigm has enabled powerful robotic control by leveraging vision-language models, but its reliance on large-scale, high-quality robot data limits…