RL towards General Intelligence RL that Reinvents Human Intelligence Reinforcement learning towards general intelligence Transferable RL Transfers across Dynamics Policies that generalize across different dynamics and environments Transfer across Tasks/ Rewards Policies that generalize across different task objectives Transfer across Embodiments Policies that generalize across different robot morphologies and bodies Scalable RL Scale via Architecture another without an image Scale via Sparse Reward a project with a background image Scale via Parallelization another project with an image 🎉
