Unified Diffusion VLA: Vision-Language-Action Model via Joint Discrete Denoising Diffusion Process

Open paper

Benchmark	Model	Rank	Results
robot-manipulation-on-calvin-abcd-to-d	UD-VLA	#3	Average Length: 4.64
robot-manipulation-on-simplerenv-widow-x	UD-VLA	#29	Average: 0.625
robot-manipulation-on-libero	UD-VLA	#119	Success Rate: 92.7