RLeXplore: Accelerating Research in Intrinsically-Motivated Reinforcement Learning
Abstract: Extrinsic rewards can effectively guide reinforcement learning (RL) agents in specific tasks. However, extrinsic rewards frequently fall short in complex environments due to the significant human effort needed for their design and annotation. This limitation underscores the necessity for intrinsic rewards, which offer auxiliary and dense signals and can enable agents to learn in an unsupervised manner. Although various intrinsic reward formulations have been proposed, their implementation and optimization details are insufficiently explored and lack standardization, thereby hindering research progress. To address this gap, we introduce RLeXplore, a unified, highly modularized, and plug-and-play framework offering reliable implementations of eight state-of-the-art intrinsic reward algorithms. Furthermore, we conduct an in-depth study that identifies critical implementation details and establishes well-justified standard practices in intrinsically-motivated RL.
Submission Length: Regular submission (no more than 12 pages of main content)
Previous TMLR Submission Url: https://openreview.net/forum?id=Y8n3Gg33mT
Changes Since Last Submission: **Desk Reject Comments**:
There should not be any figures before the title of the first section. Please fix and resubmit.
**Changes**:
The figure in the Abstract part is removed.
Assigned Action Editor: ~Aleksandra_Faust1
Submission Number: 3599
Loading