WAM-RL: World-Action Model Reinforcement Learning with Reconstruction Rewards and Online Video SFT

Add code
Jun 16, 2026

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: