The field of electronic-photonic design automation and AI-driven optimization is rapidly advancing, with a focus on developing innovative solutions to address the complexity and scalability challenges of modern electronic and photonic systems. Recent research has centered on creating automated design frameworks, leveraging machine learning and optimization techniques to improve the performance, efficiency, and reliability of these systems. Notable developments include the integration of curvy waveguides, bending, and port alignment in photonic integrated circuit design, as well as the application of computing-in-memory architectures and neural architecture search frameworks to optimize machine learning workloads. Furthermore, AI-driven optimization techniques have been successfully applied to antenna miniaturization, RF/analog circuit design, and approximate computing, demonstrating significant improvements in performance, energy efficiency, and design cycle time. Noteworthy papers in this area include LiDAR 2.0, which presents a hierarchical curvy waveguide detailed router for large-scale photonic integrated circuits, and CIM-NET, which proposes a hardware-algorithm co-design framework for computing-in-memory architectures. FALCON introduces a unified machine learning framework for fully automated layout-constrained analog circuit design, while CrossNAS presents a cross-layer neural architecture search framework for PIM systems. MenTeR and A Unified Framework for Mapping and Synthesis of Approximate R-Blocks CGRAs also demonstrate innovative approaches to RF/analog circuit design and approximate computing.