TwinHead-RL is a research project on curiosity-driven deep RL. We use a multi-head agent that learns both policy and future state prediction, using prediction error as a dynamic intrinsic reward for efficient exploration and representation learning.
-
Notifications
You must be signed in to change notification settings - Fork 0
TwinHead-RL is a research project on curiosity-driven deep RL. We use a multi-head agent that learns both policy and future state prediction, using prediction error as a dynamic intrinsic reward for efficient exploration and representation learning.
License
MorningStarTM/TwinHead-RL
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
TwinHead-RL is a research project on curiosity-driven deep RL. We use a multi-head agent that learns both policy and future state prediction, using prediction error as a dynamic intrinsic reward for efficient exploration and representation learning.
Topics
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published