Toward Efficient Permutation for Hierarchical N:M Sparsity on GPUs

Re-assign community