1 paper · 1 filter
Leonard Storcks, Tobias Buck
We present two promising applications of training machine learning models inside a differentiable astrophysical (magneto)hydrodynamics simulator. First, we address the problem of s…