The authors study offline reinforcement learning, where a policy must be learned from a fixed dataset without further interaction with the environment.
Advantage guidance makes supervised policy training competitive with offline RL
A conditional flow-matching policy and contrastive sampling method reportedly outperform strong offline RL baselines across 26 difficult control tasks.
Top University
Zhiyuan Zhou · Sergey Levine
UC Berkeley
Research Digest··2 min read
Zhou and Levine present Advantage Guidance, a policy-learning method designed to retain supervised learning’s stability while improving beyond the behavior in an offline dataset.
Why this paper
From UC Berkeley
In one line
Advantage Guidance trains a flow-matching policy conditioned on advantages and uses contrastive sampling guidance to outperform prior RL methods on 26 long-horizon, sparse reward tasks.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ·No benchmark numbers found
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§