Proxy Policy Steering: Test-Time Robot Policy Adaptation
Loading video
Loading videoProxy Policy Steering steers a frozen diffusion policy at inference time with a calibrated velocity residual from two lightweight proxies, raising average success by 53% across 12 manipulation tasks without modifying the base model.
Proxy Policy SteeringRobot Policy AdaptationDiffusion policiesInference-Time AdaptationManipulationImitation LearningCornell Universityπ0.5Robot Manipulation
Category: arm
Author: @LeoKharon
Date: 2026-09-26T00:00:00
Duration: 42.799s
Reference: https://arxiv.org/abs/2609.09148
