Disentangling continuous and discrete linguistic signals in transformer-based sentence embeddings

Open in new window