Enabling robust offline active learning for machine learning potentials using simple physics-based priors

Muhammed Shuaibi, Saurabh Sivakumar, Rui Qi Chen, Zachary W. Ulissi