The Goldilocks Paradigm: Comparing Classical Machine Learning, Large Language Models, and Few-Shot Learning for Drug Discovery Applications
Published in Nature Communications, 2024
Paper on comparing the performance of classical ML, few-shot learning and transformer-based LLM (MolBART) at varying dataset sizes
Recommended citation: Snyder, S. et. al. (2024). "The Goldilocks Paradigm: Comparing Classical Machine Learning, Large Language Models, and Few-Shot Learning for Drug Discovery Applications." Nature Communications. 1(3).
