Discovery Papers: Validated Empirical Findings Using AI

Papers whose primary contribution is a new empirical finding (material, molecule, or phenomenon) confirmed by independent experiment or simulation.
Molecular Representations
SELFIES robustness demonstration

Invalid SMILES Benefit Chemical Language Models: A Study

A 2024 Nature Machine Intelligence paper providing causal evidence that invalid SMILES generation improves chemical language model performance by filtering low-likelihood samples, while validity constraints (as in SELFIES) introduce structural biases that impair distribution learning.