R. Xu, “From Dense to Sparse: Contrastive Pruning for Better Pre-trained Language Model Compression”, AAAI, vol. 36, no. 10, pp. 11547–11555, Jun. 2022.