ResearchBenchmarksHuman Evaluation on HH datasetHuman Evaluation on HH datasetFollow59Win RateRRHF_DP-2.3613.5729.545.43Apr 11, 2023Evaluation resultsMethodMethodLinksWin RateTie RateLoss RateRRHF_DPrho=Alpaca, compared_a...rho=Alpaca, compared_against=Good responses2023.04593011RRHF_DPrho=Alpaca, compared_a...rho=Alpaca, compared_against=PPO2023.04274825RRHF_DPrho=Alpaca, compared_a...rho=Alpaca, compared_against=RRHF_IP-22023.0409010