Session topics will include "how learning language model systems are built (including the risks of biases and hallucinations) ...