Automatic differentiation isn't just for machine learning, it's useful in several other applications where you have an optimization problem built of many variab

Automatic differentiation isn't just for machine learning, it's useful in several other applications where you have an optimization problem built of many variables interacting in differentiable ways. One possible guess could be structure-from- motion or SLAM, where the graph is often dynamic and several thousands of nodes isn't uncommon (the nodes would be things such as camera poses and 3D landmarks for example). However, in that case there are other frameworks that are built with that scenario in mind (Ceres, GTSAM), and would probably be better baselines.