MoxE: Mixture of xLSTM Experts with Entropy-Aware Routing for Efficient Language Modeling

Open in new window