Training-free Generation of Temporally Consistent Rewards from VLMs

Published in ICCV, 2025