Distilling Semantic Concept Embeddings from Contrastively Fine-Tuned Language Models

Open in new window