Dong, Y. et al. 2026. EnViT: Enhancing the Performance of Early-Exit Vision Transformers via Exit-Aware Structured Dropout-Enabled Self-Distillation. Proceedings of the AAAI Conference on Artificial Intelligence. 40, 25 (Mar. 2026), 20852–20860. DOI:https://doi.org/10.1609/aaai.v40i25.39225.