arXiv Robotics — research abstracts· Wenyan Yang, Arsenii Mustafin, Dominik Baumann, Joni Pajarinen, Simone Parisi·· 12 hours agoEditorial score42
SUN: Enhancing Goal Selection in Reinforcement Learning
SUN: Reaching for Novelty in Reinforcement Learning
Summary
SUN integrates novelty and reachability in goal selection, improving performance in complex environments. It outperforms existing methods in novel and challenging scenarios.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
Source：arXiv Robotics — research abstracts · arxiv.org