List your company
ResearchFormerly WizWand

Forward PDE solving on Poisson

0.038Relative L2 ErrorDi-BiLPS
-2.1460812.5964627.33942.08154May 9, 2026May 15, 2026May 22, 2026May 28, 2026Jun 4, 2026Jun 10, 2026Jun 17, 2026
Updated 4mo ago

Evaluation results

MethodLinks
Di-BiLPS2026.050.038—19.69
DiffPDE2026.050.045—255.37
Spatio-Temporal MeanFlowSolver Category=Ours2026.050.18——
FNOSolver Category=Neural operator2026.050.23——
DiffusionPDESolver Category=Generative PDE2026.050.76——
FNO2026.051.009——
PINO2026.051.071——
PINNs2026.051.281——
HQNNParameter value (λ)=12026.061.45——
DeepONet2026.051.555——
PINOSolver Category=Neural operator2026.051.58——
Rectified FlowSolver Category=Generative PDE2026.051.58——
PINNParameter value (λ)=52026.061.66——
HQNNParameter value (λ)=62026.061.76——
PINNParameter value (λ)=62026.062.81——
PI-HPQNNParameter value (λ)=62026.063.05——
HQNNParameter value (λ)=52026.063.18——
PI-HPQNNParameter value (λ)=12026.063.28——
FPINNParameter value (λ)=52026.064.61——
DeepONetSolver Category=Neural operator2026.055.46——
PI-HPQNNParameter value (λ)=52026.066.11——
FPINNParameter value (λ)=62026.069——
FPINNParameter value (λ)=12026.069.4——
PINNParameter value (λ)=12026.069.59——
PBFMSolver Category=Generative PDE2026.0528.56——
PINNSolver Category=Physics-informed coordinate-based2026.0532.01——
I-PINNSolver Category=Physics-informed coordinate-based2026.0554.64——
DeepONetObservation mode=Sparse (5/128 points), Training setup=Full observation2024.06—155.5—
DiffusionPDEObservation mode=Sparse (5/128 points)2024.06—4.5—
DiffusionPDEnumber of steps=1000, number of observations=5002026.05—0.1532—
FM4PDEnumber of steps=1000, number of observations=5002026.05—0.0302—
FNOObservation mode=Sparse (5/128 points), Training setup=Full observation2024.06—100.9—
PINNsObservation mode=Sparse (5/128 points), Training setup=Full observation2024.06—128.1—
PINOObservation mode=Sparse (5/128 points), Training setup=Full observation2024.06—107.1—