Yftah Ziser

Latest

• Self-Improving World Modelling with Latent Actions
• Bootstrapping Action-Grounded Visual Dynamics in Unified Vision-Language Models