AI Paper Atlas

Research map / Learning theory and algorithms / Matrix and tensor methods

What if Neural Networks had SVDs?

Alexander Mathiasen, Frederik Hvilshøj, Jakob Rødsgaard Jørgensen, Anshul Nasery, Davide Mottin

NeurIPS 2020 · Learning theory and algorithms › Matrix and tensor methods · approach: Geometry of low-rank and SPD structure · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Group Householder matrices into WY products to reduce sequential operationsOrthogonal gradient descent with Householder parameterization · Orthogonal gradient descentrunning time: up to 27.1× faster than the sequential algorithmImproved
Use FastH for SVD-reparameterized matrix operationsSVD reparameterization · Matrix inversion, d=768speedup over standard inverse: 2.7×Improved
Use FastH for SVD-reparameterized matrix operationsSVD reparameterization · Cayley map, matrix exponential, and determinant, d=768speedup over standard method: 3.1×; 4.1×; 3.5×Improved
Use the sequential Householder algorithm instead of FastHSVD reparameterization · Matrix operationsrunning timeHurt
Use FastH rather than the prior parallel Householder algorithmOrthogonal gradient descent · Orthogonal gradient descentrelative speedup: up to 6.2×Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy