22 citations · 120 across the 9 of their papers we have counts for
3 papers · 1 filter
Batched Second-Order Adjoint Sensitivity for Reduced Space Methods
François Pacaud, Michel Schanen, Daniel Adrian Maldonado +4
This paper presents an efficient method for extracting the second-order sensitivities from a system of implicit nonlinear equations on upcoming graphical processing units (GPU) dom…
Instead of Rewriting Foreign Code for Machine Learning, Automatically Synthesize Fast Gradients
William S. Moses, Valentin Churavy
Applying differentiable programming techniques and machine learning algorithms to foreign programs requires developers to either rewrite their code in a machine learning framework,…
Dynamic Automatic Differentiation of GPU Broadcast Kernels
Jarrett Revels, Tim Besard, Valentin Churavy +2
We show how forward-mode automatic differentiation (AD) can be employed within larger reverse-mode computations to dynamically differentiate broadcast operations in a GPU-friendly…