tool · first seen 4 Oct 2026
fused kernel for loss and gradient
Also: Pallas kernels
Custom accelerator kernel that computes loss and gradients on-chip without materializing intermediate tensors in high-bandwidth memory.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential RecommendationPapers · 4 Oct 2026 · Yaoyiran Li, Haowen Ning, Mohamed Hammad
Latest sources
- Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential RecommendationPapers · 4 Oct 2026 · Yaoyiran Li, Haowen Ning, Mohamed Hammad