This mini-review and tutorial focuses on the fundamentals of heat transport, the relevant molecular dynamics (MD) simulation methods, and the applications of machine-learned potentials (MLPs) in MD simulations of heat transport. The authors provide a step-by-step guide on developing MLPs, particularly neuroevolution potentials (NEPs), for highly efficient and predictive heat transport simulations using the GPUMD package. NEPs, based on artificial neural networks and trained using a separable natural evolution strategy, offer superior accuracy and computational efficiency compared to traditional empirical potentials. The tutorial covers the construction of NEPs, including the selection of descriptors, training algorithms, and hyperparameter tuning. It also demonstrates the application of NEPs in various MD methods, such as equilibrium molecular dynamics (EMD), nonequilibrium molecular dynamics (NEMD), homogeneous nonequilibrium molecular dynamics (HNEMD), and spectral decomposition. The authors illustrate these concepts with a detailed example using crystalline silicon, showing how to train NEPs, validate them, and apply them to calculate thermal conductivity and phonon dispersion relations. The tutorial emphasizes the importance of NEPs in enhancing the accuracy and efficiency of MD simulations for heat transport studies, particularly for complex materials and systems.This mini-review and tutorial focuses on the fundamentals of heat transport, the relevant molecular dynamics (MD) simulation methods, and the applications of machine-learned potentials (MLPs) in MD simulations of heat transport. The authors provide a step-by-step guide on developing MLPs, particularly neuroevolution potentials (NEPs), for highly efficient and predictive heat transport simulations using the GPUMD package. NEPs, based on artificial neural networks and trained using a separable natural evolution strategy, offer superior accuracy and computational efficiency compared to traditional empirical potentials. The tutorial covers the construction of NEPs, including the selection of descriptors, training algorithms, and hyperparameter tuning. It also demonstrates the application of NEPs in various MD methods, such as equilibrium molecular dynamics (EMD), nonequilibrium molecular dynamics (NEMD), homogeneous nonequilibrium molecular dynamics (HNEMD), and spectral decomposition. The authors illustrate these concepts with a detailed example using crystalline silicon, showing how to train NEPs, validate them, and apply them to calculate thermal conductivity and phonon dispersion relations. The tutorial emphasizes the importance of NEPs in enhancing the accuracy and efficiency of MD simulations for heat transport studies, particularly for complex materials and systems.