The field of autonomous driving is rapidly advancing, with a focus on improving safety, efficiency, and scalability. Recent developments have centered around addressing the challenges of multimodal trajectory planning, incorporating constraints and safety considerations, and enhancing the robustness of end-to-end driving models. Notably, researchers have proposed innovative frameworks that leverage constrained flow matching, online world model distillation, and graph neural networks to accelerate time-optimal trajectory planning. Additionally, there has been a push towards developing more realistic and diverse simulation environments to train and evaluate autonomous driving models. Some papers have made significant contributions to the field, including the proposal of GuideFlow, which achieved state-of-the-art performance on major driving benchmarks, and AD-R1, which introduced a framework for post-training policy refinement using an Impartial World Model. WPT and Map-World also demonstrated impressive results, with WPT achieving a 0.11 collision rate and Map-World matching anchor-based approaches on the NAVSIM benchmark. Overall, the field is moving towards more robust, efficient, and scalable solutions for autonomous driving, with a strong emphasis on safety and real-world applicability.