Do you guys have any question?
@aran Does the "M" in "LLM" refer to a trained latent space representation or its architecture? Or both?