CLMay 24, 2023

Contrastive Learning of Sentence Embeddings from Scratch

arXiv:2305.15077v2142 citations
Originality Incremental advance
AI Analysis

This work addresses data acquisition issues in training sentence embeddings for NLP applications, offering a novel synthesis-based approach that is incremental but effective.

The paper tackles the challenge of acquiring data for contrastive learning of sentence embeddings by proposing SynCSE, a framework that uses large language models to synthesize data, either from unlabeled sentences or from scratch. Results show that SynCSE outperforms unsupervised baselines and achieves performance comparable to supervised models in most settings, with SynCSE-partial matching supervised models in sentence similarity and reranking tasks.

Contrastive learning has been the dominant approach to train state-of-the-art sentence embeddings. Previous studies have typically learned sentence embeddings either through the use of human-annotated natural language inference (NLI) data or via large-scale unlabeled sentences in an unsupervised manner. However, even in the case of unlabeled data, their acquisition presents challenges in certain domains due to various reasons. To address these issues, we present SynCSE, a contrastive learning framework that trains sentence embeddings with synthesized data. Specifically, we explore utilizing large language models to synthesize the required data samples for contrastive learning, including (1) producing positive and negative annotations given unlabeled sentences (SynCSE-partial), and (2) generating sentences along with their corresponding annotations from scratch (SynCSE-scratch). Experimental results on sentence similarity and reranking tasks indicate that both SynCSE-partial and SynCSE-scratch greatly outperform unsupervised baselines, and SynCSE-partial even achieves comparable performance to the supervised models in most settings.

Code Implementations2 repos
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes