Article

JevVLA: A Jev-Inspired Energy-Based Adaptation Framework for VLA Policies

Researchers propose JevVLA, a lightweight energy-based adaptation framework for vision-language-action (VLA) policies that, with the base policy frozen, learns a differentiable scalar energy function used both for comparing discrete candidates and for correcting or generating continuous actions.

Part of the event: JevVLA: An Energy-Based Adaptation Framework for Frozen Foundation Policies

This page provides a summary and source link only.