Jul 2026
Temporal Policy: History-Initialized Action Generation for Robotic Learning from Demonstration
Temporal Policy is introduced, a generative framework based on stochastic interpolants that formulates action generation as a temporally coupled transport problem and bypasses the computational bottleneck of independent Gaussian priors, helping enable high-frequency, closed-loop control.
Dylan Miller, Martin Jägersand
· arXiv.org · 0 citations