Hugo Lewi Hammer Dr.

New regularization technique boosts AI model interpretability and performance

Researchers have introduced Local Fidelity Regularization (LFR), a new method to enhance the interpretability and predictive performance of Mesomorphic Neural Networks (IMNs). LFR addresses a vulnerability in existing I…