analiza aiqbench
This arXiv paper proposes DEFlow, a reinforcement learning method that samples from expressive flow-based policies using implicit Euler integration, with gradients computed by implicit differentiation rather than through solver iterations. The reported gains are about 10% in successful coverage on Multi-Goal docking tasks and the best average return on four of five manipulation tasks, with broad competitiveness on MuJoCo benchmarks. Readers should note these are simulated benchmark results, and the method trades fewer sampling steps for extra nonlinear and linear solves at training and inference time.
DEFlow: Deep Equilibrium Flow Maximum-Entropy Reinforcement Learning
arXiv:2610.16597v1 Announce Type: new 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 system. 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 mean 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
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 system. 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 mean 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
Machine Learning (cs.LG); Robotics (cs.RO)
arXiv:2610.16597 [cs.LG]
(or arXiv:2610.16597v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.16597 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 234 cuvinte
Subiecte conexe
NVIDIA Vera CPU: Olympus Cores Built for Maximum Single-Thread Performance in Agentic AI
NVIDIA · 21 iul. 2026
MGD: Moment Guided Diffusion for Maximum Entropy Generation
arXiv · 9 sept. 2026
Perspectives on Tsallis Statistics for Artificial Intelligence
arXiv · 5 aug. 2026
Drift, Then Spread: Learning One-Step Generative Policies for Online Reinforcement Learning
arXiv · 14 oct. 2026