Inferring for Dominant Modes from X-Ray Images in OMEGA Implosions

Low-mode core asymmetries have been observed in inertial confinement fusion (ICF) implosion experiments conducted at the OMEGA laser facility [K. M. Woo et al. Physics of Plasmas 29, 082705 (2022)]. We present a novel approach for analyzing the modal spectrum of ICF implosions at OMEGA, using x-ray diagnostics using a convolution neural network (CNN). The CNN model has been trained using an extensive database of ICF simulations performed with the radiation-hydrodynamic code DEC3D postprocessed with Spect3D. Multiple hyperparameters and CNN architectures were explored to optimize model performance. The effectiveness of the CNN model was evaluated by comparing its predictions for low modes against other approaches. We will present CNN predictions and examine the modal behavior using the analysis of the singular values of the weight matrix.