The field of thermal analysis and modeling is experiencing significant advancements, driven by the increasing importance of thermal reliability in modern integrated circuits and other engineering applications. Researchers are exploring innovative approaches to improve the accuracy and efficiency of thermal analysis, including the use of generative AI, implicit physics priors, and physics-guided neural frameworks. These methods aim to address the challenges of traditional simulation-based approaches, such as high computational costs and limited scalability. Notable papers in this area include: 2D-ThermAl, which proposes a physics-informed generative AI framework for thermal analysis of circuits, achieving high accuracy and speed. Learning to Reconstruct Temperature Field from Sparse Observations with Implicit Physics Priors, which introduces an implicit physics-guided temperature field reconstruction framework, demonstrating state-of-the-art reconstruction accuracy and generalization capability. Modeling and Inverse Identification of Interfacial Heat Conduction in Finite Layer and Semi-Infinite Substrate Systems via a Physics-Guided Neural Framework, which presents a physics-guided Transformer architecture for interface-dominated diffusion problems, enabling coherent temperature field resolution and reliable identification of unknown thermal properties. Physics-Driven Learning Framework for Tomographic Tactile Sensing, which develops a physics-driven deep reconstruction framework for electrical impedance tomography, improving physical plausibility and generalization in tomographic tactile sensing.