Abstract page for arXiv paper 2608.26233: Pruning Binarized Neural Networks: A Dedicated Framework and Globally Weighted Algorithms