HASTE: Hardware-Aware Dynamic Sparse Training for Large Output Spaces 事件

HASTE: Hardware-Aware Dynamic Sparse Training for Large Output Spaces arXiv:2606.01117v1 Announce Type: cross Abstract: Extreme multi-label classification (XMC) involves learning models over large output spaces with millions of labels, making the output layer a memory-compute bottleneck. While sparsity-based methods reduce arithmetic complexity, they often fail to yield proportional speedups due to irregular memory access, poor hardware utilization, or reliance on auxiliary architectural compon