1 code implementation • 9 Nov 2023 • Felipe Urrutia, Cristian Buc, Valentin Barriere
We show that the LLM normally struggles for the main task using in-context learning, but can handle these easiest subtasks and produce useful weak labels to train a BERT.
no code implementations • 21 Apr 2023 • Felipe Urrutia, Roberto Araya
We found that LLMs perform worse than MLs in detecting incoherent answers.