analiza aiqbench
This research advances robot control by replacing likelihood-based policy gradients with flow matching methods, enabling more expressive policies for complex tasks like legged locomotion and humanoid manipulation. The approach successfully improves sim-to-real transfer and exploration capabilities without relying on differentiable action likelihoods, addressing a key limitation in current reinforcement learning frameworks for robotics.
Flow Policy Gradients for Robot Control
arXiv:2602.02481v2 Announce Type: replace-cross Abstract: Likelihood-based policy gradient methods are the dominant approach for training robot control policies from rewards. These methods rely on differentiable action likelihoods, which constrain policy outputs to simple distributions like Gaussians. In this work, we show how flow matching policy gradients, a recent framework that bypasses likelihood computation, can be made effective for training and finetuning more expressive policies in challenging robot control settings. We introduce an improved objective that enables success in legged locomotion, humanoid motion tracking, and manipulation tasks, as well as robust sim-to-real transfer on two humanoid robots. We then present ablations and analysis on training dynamics. Results show how policies can exploit the flow representation for exploration when training from scratch, as well as improved finetuning robustness over baselines.
Authors:Brent Yi, Hongsuk Choi, Himanshu Gaurav Singh, Xiaoyu Huang, Takara E. Truong, Carmelo Sferrazza, Yi Ma, Rocky Duan, Pieter Abbeel, Guanya Shi, Karen Liu, Angjoo Kanazawa
Abstract:Likelihood-based policy gradient methods are the dominant approach for training robot control policies from rewards. These methods rely on differentiable action likelihoods, which constrain policy outputs to simple distributions like Gaussians. In this work, we show how flow matching policy gradients, a recent framework that bypasses likelihood computation, can be made effective for training and finetuning more expressive policies in challenging robot control settings. We introduce an improved objective that enables success in legged locomotion, humanoid motion tracking, and manipulation tasks, as well as robust sim-to-real transfer on two humanoid robots. We then present ablations and analysis on training dynamics. Results show how policies can exploit the flow representation for exploration when training from scratch, as well as improved finetuning robustness over baselines.
Robotics (cs.RO); Artificial Intelligence (cs.AI)
arXiv:2602.02481 [cs.RO]
(or arXiv:2602.02481v2 [cs.RO] for this version)
Text integral sub licența deschisă a publicației · 162 cuvinte
Subiecte conexe
Micro-Swarm Locomotion Optimization in Dynamic Flow using Multi-Objective Multi-Agent Reinforcement Learning
arXiv · 16 iun. 2026
Q-VGM: Q-Value-Gradient Matching for Off-Policy Reinforcement Learning of Flow-Matching VLA
arXiv · 21 iul. 2026
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · 17 sept. 2026
Reinforcement Learning for Flow-Matching Policies with Density Transport
arXiv · 9 iun. 2026