Dataset Reset Policy Optimization (DR-PO): A Machine Learning Algorithm that Exploits a Generative Model’s Ability to Reset from Offline Data to Enhance RLHF from Preference-based Feedback
admin
0
0
Dataset Reset Policy Optimization (DR-PO): A Machine Learning Algorithm that Exploits a Generative Model’s Ability to Reset from Offline Data to Enhance RLHF from Preference-based Feedback
