The field of robot navigation and planning is moving towards more robust and efficient methods for handling dynamic and uncertain environments. Researchers are focusing on developing frameworks that can accurately model multi-scale temporal dependencies and balance exploration and exploitation during training. This has led to significant improvements in success rates and path quality. Notable advancements include the integration of uncertainty into decision-making processes and the development of hybrid model- and learning-based approaches for traversability estimation. These innovations have the potential to enhance the safety and efficiency of robot navigation in real-world environments. Noteworthy papers include: GundamQ, which proposes a multi-scale spatiotemporal Q-network for robotic path planning, achieving a 15.3% improvement in success rate and a 21.7% increase in overall path quality. NAMOUnc, which introduces a novel framework for navigation among movable obstacles with decision making on uncertainty interval, demonstrating significant improvements over existing frameworks. NavMoE, which presents a hierarchical and modular approach for traversability estimation and local navigation, offering improved efficiency and performance balance across different domains. Bridging Perception and Planning, which proposes a differentiable framework for end-to-end planning for Signal Temporal Logic tasks, outperforming single-expert baselines in STL satisfaction and trajectory feasibility.