POLLU

The 'POLLU' dataset is a real-world stiff atmospheric chemistry benchmark containing 25 parameters, used to evaluate the performance of the Inverse Neural Operator in recovering hidden ODE parameters from sparse observations.

Papers using POLLU (3)

Inverse Neural Operator for ODE Parameter Optimization2026MENO: Hybrid Matrix Exponential-based Neural Operator for Stiff ODEs. Application to Thermochemical Kinetics2025Learning stiff chemical kinetics using extended deep neural operators2023 · 56 cites