RLHF is good at doing what humans want the model to do, or at least to state it more precisely doing what humans who look at the model for…
Not checkable An opinion, value judgement or statement with nothing to check.
Not checkable: An opinion, value judgement or statement with nothing to check. Not scored.
Cite this claim
Dario Amodei, “RLHF is good at doing what humans want the model to do, or at least to state it more precisely doing what humans who look at the model for…” (said 11 November 2024). Verdict: not checkable. Reliability API, claim c_bc681d09b15c, method 0.1.0+c876030c, dataset release v1.1.2. https://epistacklabs.com/claims/c_bc681d09b15c
This claim is in dataset release v1.1.2, with the same verdict.
Reply to this claim
Verdicts are made automatically; how verdicts are made.