On the class separability of contextual embeddings representations - or "The classifier does not matter when the (text) representation is so good!".

@article{andrade-2023-class, author = {Cl\'{a}udio Mois\'{e}s Valiense de Andrade and Fabiano Bel\'{e}m and Washington Cunha and Celso Fran\c{c}a and Felipe Viegas and Leonardo Rocha and Marcos Andr\'{e} Gon\c{c}alves}, title = {{On the class separability of contextual embeddings representations - or "The classifier does not matter when the (text) representation is so good!"}}, journal = {Information Processing and Management}, volume = {60}, number = {4}, pages = {103336}, year = {2023}, url = {https://doi.org/10.1016/J.IPM.2023.103336}, doi = {https://doi.org/10.1016/J.IPM.2023.103336} }