Recovers hidden ODE parameters from sparse data with a 487x speedup over gradient-based methods.
March 13, 2026
Original Paper
Inverse Neural Operator for ODE Parameter Optimization
arXiv · 2603.11854
The Takeaway
The Inverse Neural Operator (INO) bypasses Jacobian instabilities in stiff regimes by using a drifting model in parameter space. This allows for near-instantaneous (0.2s) system identification in complex fields like atmospheric chemistry and gene regulation.
From the abstract
We propose the Inverse Neural Operator (INO), a two-stage framework for recovering hidden ODE parameters from sparse, partial observations. In Stage 1, a Conditional Fourier Neural Operator (C-FNO) with cross-attention learns a differentiable surrogate that reconstructs full ODE trajectories from arbitrary sparse inputs, suppressing high-frequency artifacts via spectral regularization. In Stage 2, an Amortized Drifting Model (ADM) learns a kernel-weighted velocity field in parameter space, trans