Last released Feb 1, 2026
A Triton implementation of greedy submodular algorithms using fused GPU kernels
Supported by