New article in A&A 📃

My new article Bypassing scaling relations with Lagrangian Deep Learning and Simulation-based inference has been accepted in A&A and will be published soon ! It presents a new method to model cluster number counts, using cosmological simulations and ML.

We developped a fast field-level emulator powered by JaxPM and Lagrangian Deep Learning to produce cluster catalogs. The model is calibrated on CAMELS simulations and allows We then used implicit inference to obtain contours on the cosmology and the astrophysics. Here below are mock X-ray maps generated by the pipeline

Scaling relation between Xray flux and mass for cluster catalogs, as in the original CAMELS simulations (blue) and in the accelerated LDL pipeline (red)