Offline Goal Conditioned Reinforcement Learning
Offline goal-conditioned reinforcement learning (GCRL) focuses on training robots to achieve various goals using only pre-collected data, eliminating the need for extensive online interaction. Current research emphasizes improving the learning of optimal policies from often suboptimal or sparse offline datasets, exploring techniques like metric learning to better represent value functions, conditional diffusion models for trajectory generation, and novel reward shaping methods. These advancements are significant because they enable the development of more robust and versatile robotic agents capable of performing complex tasks in real-world scenarios with limited data and without extensive manual reward engineering.
Papers
October 26, 2024
March 15, 2024
March 4, 2024
February 16, 2024
February 11, 2024
November 3, 2023
October 30, 2023
July 22, 2023
March 16, 2023
February 17, 2023
February 15, 2023
February 7, 2023
January 5, 2023
August 28, 2022