Interpretable lightweight transformer via unrolling of learned graph smoothness priors

Published in Advances in Neural Information Processing Systems 37 (NeurIPS 2024), 2024