The field of VLSI and AMS circuit analysis is experiencing significant growth, driven by the need for more accurate and efficient methods for analyzing complex circuits. Recent developments have focused on leveraging graph neural networks and contrastive learning to improve the accuracy of parasitic estimation and capacitance prediction. These advancements have the potential to greatly reduce the time and cost associated with circuit design and simulation, enabling the development of more complex and efficient systems. Notably, innovative approaches such as graph representation learning and few-shot learning are being applied to address the challenges of scarce data and heterogeneous circuit graphs. Overall, the field is moving towards more robust and transferable models that can handle the diversity of circuit implementations. Some particularly noteworthy papers include: GATMesh, which achieves high accuracy with average delay error of 5.27ps on unseen benchmarks, while achieving speed-ups of 47146x over multi-threaded SPICE simulation. CircuitGCL, which outperforms all state-of-the-art methods, with the R^2 improvement of 33.64% ~ 44.20% for edge regression and F1-score gain of 0.9x ~ 2.1x for node classification. CircuitGPS, which improves the accuracy of coupling existence by at least 20% and reduces the MAE of capacitance estimation by at least 0.067 compared to existing methods.