AI Paper Atlas

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

Multilevel CNNs for Parametric PDEs

Cosmas Heiß, Ingo Gühring, Martin Eigel

JMLR 2023 · Probabilistic and scientific modeling › Differential equations and PDEs · approach: Neural PDE solvers · 8 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
train ML-Net with halved sample count per finer level instead of the full 1000 samples (ablation)ML-Net · uniform p=100MR H1 error: 2.27e-4 -> 2.86e-4No clear effect
train ML-Net with halved sample count per finer level instead of the full 1000 samples (ablation)ML-Net · log-normal p=100MR H1 error: 2.64e-4 -> 4.05e-4No clear effect
use ML-Net on decaying dataset instead of UNetSeq on 232 full-resolution samplesML-Net · uniform p=100MR H1 error: 1.14e-3 -> 2.86e-4Improved
use ML-Net on decaying dataset instead of UNetSeq on 232 full-resolution samplesML-Net · log-normal p=100MR H1 error: 5.92e-3 -> 4.05e-4Improved
reduce UNetSeq training set from 1000 to 232 full-resolution samples (ablation)UNetSeq · log-normal p=100MR H1 error: 2.82e-3 -> 5.92e-3Hurt
use multilevel ML-Net instead of UNetSeqML-Net · uniform p=50MR H1 error: 2.08e-3 -> 2.21e-4Improved
use multilevel ML-Net instead of UNetSeqML-Net · log-normal p=10MR H1 error: 2.83e-3 -> 2.15e-4Improved
use multilevel ML-Net instead of UNetSeqML-Net · uniform p=10MR H1 error: 9.75e-5 -> 2.24e-4Hurt
use multilevel ML-Net instead of UNetSeqML-Net · cookie variable p=128MR H1 error: 3.08e-3 -> 5.98e-3Hurt
use multilevel ML-Net instead of UNetSeqML-Net · cookie fixed p=64MR H1 error: 7.20e-4 -> 1.85e-3Hurt
increase parameter dimension p from 10 to 200ML-Net · uniformMR H1 error: 2.24e-4 -> 2.31e-4No clear effect
increase parameter dimension p from 16 to 64ML-Net · cookie fixedMR H1 error: 9.41e-4 -> 1.85e-3Hurt

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