Diffusion-QL uses conditional diffusion models as expressive policies in offline RL by coupling behavior cloning with Q-value maximization, achieving SOTA on most D4RL tasks.
Off-policy deep reinforcement learning without exploration
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2022 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning
Diffusion-QL uses conditional diffusion models as expressive policies in offline RL by coupling behavior cloning with Q-value maximization, achieving SOTA on most D4RL tasks.