Argos is a verification framework for multimodal reinforcement learning that trains models by rewarding not just correct answers, but correct answers grounded in visual and temporal evidence, using automated verification rather than human labeling. It selects the appropriate specialized tools for each answer based on what needs to be verified. Models trained with Argos show stronger spatial reasoning, far fewer visual hallucinations, more stable learning dynamics, and better performance on robotics and real-world tasks while requiring fewer training samples.
مشاهده مطلب اصلی