Borhane Blili-Hamelin
@borhane.bsky.social
Excited to have our mini paper on the ETHICS benchmark at the @neuripsconf.bsky.social #EvalEval workshop next week! We draw on moral theory, empirical research, and prompt evaluation to argue that the benchmark lacks validity. Stay tuned for future work on the practical consequences for evals.
December 2, 2024 at 4:59 PM
Excited to have our mini paper on the ETHICS benchmark at the @neuripsconf.bsky.social #EvalEval workshop next week! We draw on moral theory, empirical research, and prompt evaluation to argue that the benchmark lacks validity. Stay tuned for future work on the practical consequences for evals.