Abstract: In this work, we replace Mamba in UMamba with recent xLSTM, and surprisingly, it works well! Convolutional Neural Networks (CNNs) and Vision Transformers (ViT) have been pivotal in ...