AI Paper Atlas

Research map / Probabilistic and scientific modeling / Differential equations and PDEs

Learning Differential Equations that are Easy to Solve

Jacob Kelly, Jesse Bettencourt, Matthew J Johnson, David K. Duvenaud

NeurIPS 2020 · Probabilistic and scientific modeling › Differential equations and PDEs · approach: Neural ODE training and analysis · 51 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
Add second-total-derivative speed regularizationNeural ODE classifier · MNIST classificationNFE: 110 -> 104Improved
Add second-total-derivative speed regularizationNeural ODE classifier · MNIST classificationLoss: .0235 -> .0234Improved
Use Taylor-mode speed regularization instead of RNODE regularizationFFJORD · MINIBOONENFE: 182 -> 177Improved
Use Taylor-mode speed regularization instead of RNODE regularizationFFJORD · MINIBOONELoss: 9.77 -> 9.84Hurt
Add second-total-derivative speed regularizationFFJORD · MNIST density estimationNFE: 149 -> 92Improved
Add second-total-derivative speed regularizationFFJORD · MNIST density estimationBits/Dim: 1.033 -> 1.039Hurt
Add trajectory-derivative speed regularizationLatent ODE · PhysioNet Challenge 2012Average NFE: 281 to 90Improved
Regularize the solver-matched second derivative rather than another order (ablation)Neural ODE classifier · MNIST classification with a 2nd-order solverSpeed-performance tradeoffImproved
Regularize the solver-matched third derivative rather than another order (ablation)Neural ODE classifier · MNIST classification with a 3rd-order solverSpeed-performance tradeoffImproved
Vary regularization order with a 5th-order solver (ablation)Neural ODE classifier · MNIST classification with a 5th-order solverSpeed-performance tradeoffMixed
Increase regularization order for an adaptive-order solver (ablation)Neural ODE classifier · MNIST classification with an adaptive-order solverSpeed-performance tradeoffImproved
Use Taylor-mode AD rather than nested forward-mode differentiationHigher-order derivative computation · Neural ODE trainingWall-clock computation time: an order of magnitudeImproved

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