Sentence Embeddings

SimCSE is a contrastive learning framework for generating sentence embeddings. It utilizes an unsupervised approach, which takes an input sentence and predicts itself in contrastive objective, with only standard dropout used as noise. The authors find that dropout acts as minimal “data augmentation” of hidden representations, while removing it leads to a representation collapse. Afterwards a supervised approach is used, which incorporates annotated pairs from natural language inference datasets into the contrastive framework, by using “entailment” pairs as positives and “contradiction

Source: SimCSE: Simple Contrastive Learning of Sentence Embeddings


Paper Code Results Date Stars


Task Papers Share
Sentence 36 25.90%
Sentence Embeddings 28 20.14%
Semantic Textual Similarity 18 12.95%
Sentence Embedding 13 9.35%
Language Modelling 7 5.04%
Retrieval 6 4.32%
Semantic Similarity 4 2.88%
Question Answering 2 1.44%
Information Retrieval 2 1.44%


Component Type
🤖 No Components Found You can add them if they exist; e.g. Mask R-CNN uses RoIAlign