I can simply take a hessian of some function in multiple points hess = x -> (Zygote.hessian.(x->sum(x.^3), x)) hess(rand(2,3)) 2×3 Array{Float64,2}: 2.67795 3.8318 4.9022 0.216379 5.67439 4.96389 But I’m always getting in troubles when I’m trying to implement the same operation for a neur…

In the first example, do you mean to take the hessian w.r.t each column of the matrix? Because the hessian of a function from \mathbb{R}^2 \to \mathbb{R} should be in \mathbb{R}^{2 \times 2}. I think you perhaps meant to do this: julia> hess = x -> (Zygote.hessian(x->sum(x.^3), x)) # note that it's…

Apply hessian of neural network to array

Specific Domains Machine Learning

Sunny March 20, 2021, 2:35pm 3

This is really what I meant and what I looked for.
I appreciate your help, thank you

Topic		Replies	Views
Is there an efficient way to compute the Hessian of a NN? Machine Learning flux	11	4738	November 12, 2019
Hessian inside a Flux loss function Machine Learning question	3	807	February 26, 2021
Calculating hessian of a NN w.r.t params New to Julia question , zygote	0	348	February 24, 2021
Fast Hessian size increase Numerics differentiation , sparse	6	563	March 19, 2021
Hessian matrix of ML model General Usage flux , zygote , forwarddiff , reversediff	9	2309	April 28, 2021

Apply hessian of neural network to array

Related topics