Deep learning is becoming increasingly important in engineering and science, especially in the field of computational mechanics. It's crucial for modeling physical phenomena and solving complex problems related to material properties. Physics-Informed Neural Networks (PINNs) and similar technologies are transforming how we handle multiphysics and inverse problems, providing a viable option beyond traditional Reduced Order Models (ROM). Recent progress has brought about various deep learning methods designed for computational mechanics, which are great for analyzing fracture mechanics, materials that behave inelastically, and multiphase poroelasticity. Another important area, uncertainty quantification, blends well with machine learning, particularly when data is scarce or unreliable. Established methods such as Bayesian frameworks or Gaussian Processes have been merged with deep learning to improve optimization and manage randomness. Notable examples are Bayesian Physics-Informed Neural Networks (B-PINNs) and Deep Kernel Learning (DKL), which effectively tackle both straightforward and complex problems. For this Mini Symposium, we are looking for papers that explore deep learning in physics modelling, Bayesian and Gaussian processes, and their uses in solid mechanics. We aim to foster discussions among specialists in computational mechanics and related fields to share ideas and breakthroughs that push our knowledge and skills forward in these crucial domains.