policy

PCP-Dreamer

IAMHUT · PyTorch

or hover any field below to flag it

Overview

Name
PCP-Dreamer
Author
IAMHUT
Framework
PyTorch
License
unknown
Skill type
other
Evidence level
untested
Task description
This code implements PCP-Dreamer, a world-model reinforcement learning algorithm with Prospective Cognitive Pruning (PCP). An RSSM is trained with reconstruction and balanced KL losses, and policies are optimized in latent imagination using pathwise derivatives and lambda-returns.

Spaces

Action space
other · 0-dim · 0Hz
Observation space
  • type: other

Links

HuggingFace repo
null
Paper (arXiv)
null

Compatible robots

0+20 mentioned but not in catalog yet

No robots list PCP-Dreamer as compatible yet. Know of one? Flag it above.

Compatible environments

0

No environments list PCP-Dreamer yet.

Datasets that reference this policy

0

No datasets reference PCP-Dreamer yet.