Simulation-Efficient Cosmological Inference with Multi-Fidelity Simulation-Based Inference

Leander Thiele, Adrian E. Bayer, Naoya Takeishi

Abstract: The simulation cost for cosmological simulation-based inference can be decreased by combining simulation sets of varying fidelity. We propose an approach to such multi-fidelity inference based on feature matching and knowledge distillation. Our method results in improved posterior quality, particularly for small simulation budgets and difficult inference problems.