Flow Policy Gradients for Robot Control

Brent Yi, Hongsuk Choi, Himanshu Gaurav Singh, Xiaoyu Huang, [0.25ex] Takara E. Truong, Carmelo Sferrazza, Yi Ma, Rocky Duan, [0.25ex] Pieter Abbeel, Guanya Shi, Karen Liu, Angjoo Kanazawa, [1.5ex] Amazon FAR, [0.5ex] *Equal Contribution Work done as an intern at Amazon FAR Amazon FAR team co-lead · 2026

Read Flow Policy Gradients for Robot Control by Brent Yi, Hongsuk Choi, Himanshu Gaurav Singh, Xiaoyu Huang, [0.25ex] Takara E. Truong, Carmelo Sferrazza, Yi Ma, Rocky Duan, [0.25ex] Pieter Abbeel, Guanya Shi, Karen Liu, Angjoo Kanazawa, [1.5ex] Amazon FAR, [0.5ex] *Equal Contribution Work done as an intern at Amazon FAR Amazon FAR team co-lead (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-02481.txt.

Description

Brent Yi and 11 co-authors make flow-based policies trainable with policy gradients for hard robot control, in a 2026 arXiv paper. Standard policy gradient methods need differentiable action likelihoods, which restricts policies to simple distributions like Gaussians. Flow matching policies are more expressive but bypass likelihood computation, so the usual training math fails.

The authors introduce an improved objective that makes flow policy gradients work in practice. It succeeds on legged locomotion, humanoid motion tracking, and manipulation tasks. Policies also transferred robustly from simulation to two real humanoid robots. Ablations examine the training dynamics in detail. Flow policies proved able to exploit their representation for exploration when learning from scratch. Fine-tuning was more robust than with baselines, suggesting expressive policies need not sacrifice trainability.

Catalog record

CREATORBrent Yi, Hongsuk Choi, Himanshu Gaurav Singh, Xiaoyu Huang, [0.25ex] Takara E. Truong, Carmelo Sferrazza, Yi Ma, Rocky Duan, [0.25ex] Pieter Abbeel, Guanya Shi, Karen Liu, Angjoo Kanazawa, [1.5ex] Amazon FAR, [0.5ex] *Equal Contribution Work done as an intern at Amazon FAR Amazon FAR team co-lead
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE59 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-02481

Browse the full library · About the Tower of Babel