An Adam-style adaptive optimizer with Katyusha momentum is applied to six diffusion-policy RL methods and reported to beat or match baselines on six robotic control tasks.
Katyusha: The first direct acceleration of stochastic gradient methods
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Adaptive Diffusion Policy Optimization for Robotic Manipulation
An Adam-style adaptive optimizer with Katyusha momentum is applied to six diffusion-policy RL methods and reported to beat or match baselines on six robotic control tasks.