SyncAI.news, a Varaisys broadcasting
QAMM: Adjoint MeanFlow Matching for Few-Step Offline Reinforcement Learning
YG

Yuehu Gong, Shutong Ding, Mokai Pan, Yimiao Zhou, Jiashu Hou, Ye Shi, Yanwei Fu

· 1 min read

ResearcharXiv cs.LG

QAMM: Adjoint MeanFlow Matching for Few-Step Offline Reinforcement Learning

arXiv:2609.34497v1 Announce Type: new Abstract: Flow policies can model rich action distributions, but their iterative sampling limits decision speed. Adjoint matching uses the critic's action gradient to improve a flow policy without backpropagating through its sampling trajectory, yet its supervision is defined for instantaneous velocities. We propose QAMM, a method that turns the critic-derived adjoint signal into supervision for MeanFlow's average velocity. The resulting policy learns finite-interval transport directly and generates actions with few network evaluations. We derive the adjoint MeanFlow target, specify its gradient boundaries, and train it with an offline actor-critic. On ten HumanoidMaze tasks, QAMM produces effective two-call policies and achieves competitive performance against strong flow-policy baselines. These results show that adjoint-based Q optimization can be combined with average-velocity learning to obtain expressive offline policies with few-step action generation.

Original source

This story was published by arXiv cs.LG and written by Yuehu Gong, Shutong Ding, Mokai Pan, Yimiao Zhou, Jiashu Hou, Ye Shi, Yanwei Fu. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on arxiv.org

Similar News