SOURCE-LINKED INTELLIGENCE

Proxy Policy Steering

arXiv · AI, language, vision and robotics · article · Sep 8, 2026 · UTC

Generalist robot policies carry broad manipulation priors from large-scale data, but specializing them to a new task remains the deployment bottleneck. This requires eliciting task-specific behavior from limited demonstrations without degrading their broad capabilities. We introduce Proxy Policy Steering (PPS), an inference-time adaptation method that resolves this challenge by training two lightweight proxy policies whose calibrated velocity-space difference steers the frozen base sampler. A reference proxy models the frozen base's behavior on target-task observations, and a task proxy, initi

Read original source ↗ Open in workspace

recordType

paper

region

Global