2 papers
cs.PL2025
Latency Based Tiling
Jack Cashman
Latency Based Tiling provides a systems based approach to deriving approximate tiling solution that maximizes locality while maintaining a fast compile time. The method uses triang…
cs.PF2025
Tiny-QMoE
Jack Cashman, Jiaqi Nie
The QMoE model provides a practical approach for compression of massive Mixture-of-Experts (MoE) models. QMoE offers a solution geared towards memory limitations that often reach t…