February 21, 2026 Artificial Intelligence Computation and Language Machine Learning Transformers Unleashed: From Interpretability to Efficiency and Beyond Latest 18 papers on transformer models: Feb. 21, 2026