Official release of PRIMO R1, a 7B video MLLM for robotic process reasoning featuring RL-optimized models, SFT/RL datasets, and cross-domain benchmark
-
LeonOverload/PRIMO-COT-SFT-7B
Video-Text-to-Text • 849k • Updated • 36 • 1 -
LeonOverload/PRIMO-R1-7B
Video-Text-to-Text • 8B • Updated • 39 • 1 -
From Passive Observer to Active Critic: Reinforcement Learning Elicits Process Reasoning for Robotic Manipulation
Paper • 2603.15600 • Published • 7 -
LeonOverload/primo-bench-json
Viewer • Updated • 23.7k • 156
