Wang, Yingchao, et al. “Rethinking the Dark Knowledge and Kullback-Leibler Divergence Loss in Knowledge Distillation Under Capacity Mismatching”. Proceedings of the AAAI Conference on Artificial Intelligence, vol. 40, no. 31, Mar. 2026, pp. 26688-96, doi:10.1609/aaai.v40i31.39878.