CAINCCorporate AI Narrative CorpusPILOT / 0.8
Back to interactive corpus
CAINC-OAI-2017-001

Learning from human preferences

OpenAI · research-announcement · 2017-06-13

Short quotation / metadata

Record summary

Describes human feedback as a way to align reinforcement-learning goals with human preferences.

This page exposes metadata and links to the original source. Refer to the interactive corpus for available public text or quotation context.

Open original source