Alleviating Over-smoothing for Unsupervised Sentence Representation
This addresses a specific bottleneck in natural language processing for improving sentence embeddings, but it is incremental as it builds on existing contrastive learning methods.
The paper tackles the over-smoothing problem in unsupervised sentence representation learning, which reduces the capacity of pre-trained language models, and presents Self-Contrastive Learning (SSCL) to alleviate it, achieving superior performance improvements on Semantic Textual Similarity and Transfer datasets.
Currently, learning better unsupervised sentence representations is the pursuit of many natural language processing communities. Lots of approaches based on pre-trained language models (PLMs) and contrastive learning have achieved promising results on this task. Experimentally, we observe that the over-smoothing problem reduces the capacity of these powerful PLMs, leading to sub-optimal sentence representations. In this paper, we present a Simple method named Self-Contrastive Learning (SSCL) to alleviate this issue, which samples negatives from PLMs intermediate layers, improving the quality of the sentence representation. Our proposed method is quite simple and can be easily extended to various state-of-the-art models for performance boosting, which can be seen as a plug-and-play contrastive framework for learning unsupervised sentence representation. Extensive results prove that SSCL brings the superior performance improvements of different strong baselines (e.g., BERT and SimCSE) on Semantic Textual Similarity and Transfer datasets. Our codes are available at https://github.com/nuochenpku/SSCL.