the Evaluation is based on some ongoing research with and @sreejan_kumar and @princeton_nlp. with any luck , we'll publish one thing empirical before long. but in the necessarily mean time this was an exciting application to help make :)In either scenario, my supervisor said that I could hold a brief seminar to demonstrate GPT-4. If I try this,