Neural network framework for volatility surface approximation and calibration. Supports rough Heston/Bergomi, random grids, multi-regime architectures.
-
Updated
Jul 7, 2026 - Python
Neural network framework for volatility surface approximation and calibration. Supports rough Heston/Bergomi, random grids, multi-regime architectures.
Volatility surface modelling library
Deterministic error certificates for rough Heston prices and portfolios, combining fractional residual bounds with shared Fourier error geometry.
GPU-accelerated quantitative derivatives engine: FiLM-FNO rough volatility calibration (<1ms), multi-fidelity PDE ensemble autocall pricing (0.99 bps RMSE, SR 26-2 risk guardian), deep hedging, and Next.js/FastAPI streaming (1,126 tests).
To associate your repository with the rough-heston topic, visit your repo's landing page and select "manage topics."