The field of video generation and robotic manipulation is moving towards more personalized and controllable models. Recent developments have focused on improving the fidelity and realism of generated videos, as well as enhancing the control and flexibility of robotic systems. Notably, innovative approaches have been proposed to address the challenges of identity preservation, temporal coherence, and physical plausibility in video generation. Additionally, new frameworks have been introduced to enable more efficient and scalable control of visual concepts and robotic skills. Overall, the field is advancing towards more sophisticated and realistic models that can be applied to a wide range of applications. Noteworthy papers include: Lynx, which introduces a high-fidelity model for personalized video synthesis, and World4RL, which proposes a framework for refining pre-trained policies in robotic manipulation using diffusion-based world models. Text Slider is also notable for its efficient and plug-and-play framework for continuous concept control in image and video synthesis. PhysCtrl is another significant contribution, introducing a novel framework for physics-grounded image-to-video generation with physical parameters and force control.