CLEAR: Contrastive Learning for Sentence Representation
This work addresses the problem of learning robust sentence representations for various downstream NLP tasks, which is important for researchers and practitioners working with language models.
This paper proposes CLEAR, a novel contrastive learning framework for sentence representation that utilizes multiple sentence-level augmentation strategies. The method outperforms existing approaches on both SentEval and GLUE benchmarks.
Pre-trained language models have proven their unique powers in capturing implicit language features. However, most pre-training approaches focus on the word-level training objective, while sentence-level objectives are rarely studied. In this paper, we propose Contrastive LEArning for sentence Representation (CLEAR), which employs multiple sentence-level augmentation strategies in order to learn a noise-invariant sentence representation. These augmentations include word and span deletion, reordering, and substitution. Furthermore, we investigate the key reasons that make contrastive learning effective through numerous experiments. We observe that different sentence augmentations during pre-training lead to different performance improvements on various downstream tasks. Our approach is shown to outperform multiple existing methods on both SentEval and GLUE benchmarks.