Explore >> Select a destination
|
You are here 
|
nhigham.com | ||
| | | | |
blog.autarkaw.com
|
|
| | | | | ||
| | | | |
sookocheff.com
|
|
| | | | | In a purely functional language - like lambda calculus - programs are expressed as nested function calls. Repetition in such an environment requires that nesting of function calls continues until some condition is met. During the repetition, each function passes its result to the next function in the nested chain and this repetition is completed when a test for some condition passes. The repetitive behaviour I've just described is recursion: | |
| | | | |
azdavis.net
|
|
| | | | | Various varieties of function in programming languages. | |
| | | | |
iclr-blogposts.github.io
|
|
| | | The product between the Hessian of a function and a vector, the Hessian-vector product (HVP), is a fundamental quantity to study the variation of a function. It is ubiquitous in traditional optimization and machine learning. However, the computation of HVPs is often considered prohibitive in the context of deep learning, driving practitioners to use proxy quantities to evaluate the loss geometry. Standard automatic differentiation theory predicts that the computational complexity of an HVP is of the same order of magnitude as the complexity of computing a gradient. The goal of this blog post is to provide a practical counterpart to this theoretical result, showing that modern automatic differentiation frameworks, JAX and PyTorch, allow for efficient computat... | ||
