Interpretable and Explainable Surrogate Modeling for Simulations: A State-of-the-Art Survey and Perspectives on Explainable AI for Decision-Making

Develop a toolkit for wrapping opaque scientific simulators with interpretable surrogate models. Focus on post-hoc explanation modules that map physical outcomes to model features.

Suggested repo: surrogate-lens

"Make your black-box scientific simulators transparent and interpretable."

Estimated effort: 60h