DEFlow: Deep Equilibrium Flow for Maximum-Entropy Reinforcement Learning
DEFlow: Deep Equilibrium Flow Maximum-Entropy Reinforcement Learning
DEFlow improves policy sampling with implicit integration, achieving 10% better coverage in docking tasks and top returns in manipulation. It balances diversity and precision with extra computational cost.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
manipulation tasks
5
View original evidence
n time-averaged return on four of five manipulation tasks, while remaining broadly competitive across five MuJoCo tasks. These results support implicit flow policies for tasks requiring both diversity and precision, at the cost of additional nonlinear and linear solves. Primary
Open source S4MuJoCo tasks
5
View original evidence
n time-averaged return on four of five manipulation tasks, while remaining broadly competitive across five MuJoCo tasks. These results support implicit flow policies for tasks requiring both diversity and precision, at the cost of additional nonlinear and linear solves. Primary
Open source S4Multi-Goal docking variants
4
View original evidence
ystem. We combine implicit differentiation with a discrete change-of-variables formula for the final-action density, enabling off-policy training within soft actor-critic without backpropagating through solver iterations. DEFlow achieves an average relative improvement of approximately 10% in mean successful coverage over the strongest baseline across four Multi-Goal docking variants. It also achieves the highest mea
Open source S3
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ystem. We combine implicit differentiation with a discrete change-of-variables formula for the final-action density, enabling off-policy training within soft actor-critic without backpropagating through solver iterations. DEFlow achieves an average relative improvement of approximately 10% in mean successful coverage over the strongest baseline across four Multi-Goal docking variants. It also achieves the highest mea
Open source S3
n time-averaged return on four of five manipulation tasks, while remaining broadly competitive across five MuJoCo tasks. These results support implicit flow policies for tasks requiring both diversity and precision, at the cost of additional nonlinear and linear solves. Primary
Open source S4
arXiv:2610.16597v2 Announce Type: replace-cross Abstract: Expressive flow policies can represent multiple high-value actions, but sampling them with few integration steps while retaining tractable action densities remains challenging. We introduce DEFlow, a deep equilibrium flow policy that replaces explicit Euler sampling with implicit Euler integration and solves all intermediate actions jointly as an equilibrium s
Open source S2
Source：arXiv Robotics — research abstracts · arxiv.org