vp-vla-visual-prompting-robotics

Replace monolithic VLA single-pathway decision-making with a decomposed System-2/System-1 architecture where a pretrained VLM planner identifies targets as visual prompts (crosshairs, bounding boxes) and a VLA controller executes on grounded observations, improving success rates by 5-8% on manipulation tasks. Use when spatial precision is critical, multi-st…

Independent GitHub snapshot · Automated classification · Original documentation governs