Kernel-Elastic Autoencoder for Molecular Design

arXiv:2310.08685v2 Announce Type: replace
Abstract: We introduce the Kernel-Elastic Autoencoder (KAE), a self-supervised generative model based on the transformer architecture with enhanced performance for molecular design. KAE is formulated based on two novel loss functions: modified maximum mean discrepancy and weighted reconstruction. KAE addresses the long-standing challenge of achieving valid generation and accurate reconstruction at the same time. KAE achieves remarkable diversity in molecule generation while maintaining near-perfect reconstructions on the independent testing dataset, surpassing previous molecule-generating models. KAE enables …

abstract architecture arxiv autoencoder challenge cs.lg design elastic functions generative kernel loss mean novel performance transformer transformer architecture type