deontologician(Josh Kuhn)

• deontologician 15 Mar 2023 5:14 UTC

  2 points

  1

  in reply to: hazel’s comment on: ARC tests to see if GPT-4 can es­cape hu­man con­trol; GPT-4 failed to do so

  Yeah, if the RLHF is supposed to train honesty into it, maybe it would have done worse on the task rabbit task. This really seems like a PR throwaway line rather than a legit attempt to red team the final model.