Kavli Affiliate: Jing Wang | First 5 Authors: Huazheng Wang, Daixuan Cheng, Haifeng Sun, Jingyu Wang, Qi Qi | Summary: Transformer-based pretrained language models (PLMs) have achieved great success in modern NLP. An important advantage of PLMs is good out-of-distribution (OOD) robustness. Recently, diffusion models have attracted a lot of work to apply diffusion to […]
Continue.. How Does Diffusion Influence Pretrained Language Models on Out-of-Distribution Data?