Nassima Aleb

Deep Learning for Polymer Informatics: A Critical Analysis of Representations, Architectures, and Evaluation Practices

It is argued that realizing deep learning’s full potential requires not only architectural innovation but also domain-aware representations that encode the statistical ensemble nature of polymers, evaluation protocols aligned with discovery scenarios, and physically grounded inductive biases.