PLAN-SEQ-LEARN: A Machine Learning Method that Integrates the Long-Horizon Reasoning Capabilities of Language Models with the Dexterity of Learned Reinforcement Learning RL Policies
TLDRA modular solution called PSL integrates language models with reinforcement learning for robotic tasks, achieving a success rate above 85% and outperforming existing methods in contact-rich tasks.
•4m read time
Be the first to comment.