ICLR2021对比学习（Contrastive Learning）NLP领域论文进展梳理

本文挑选了ICLR2021中NLP领域下的六篇文章进行解读，包含了文本生成、自然语言理解、预训练语言模型训练和去偏、以及文本匹配和文本检索。

1、Contrastive Learning with Adversarial Perturbations for Conditional Text Generation

2、CoDA: Contrast-enhanced and Diversity-promoting Data Augmentation for Natural Language Understanding

3、FairFil: Contrastive Neural Debiasing Method for Pretrained Text Encoders

4、Towards Robust and Efficient Contrastive Textual Representation Learning

5、Self-supervised Contrastive Zero to Few-shot Learning from Small, Long-tailed Text data

6、Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval

内容中包含的图片若涉及版权问题，请及时与我们联系删除