Training language models to follow instructions with human feedback, Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul Christiano, Jan Leike, Ryan Lowe, 2022Advances in Neural Information Processing Systems 35 (NeurIPS 2022)DOI: 10.48550/arXiv.2203.02155 - Describes the practical application of human feedback, including the collection of pairwise preferences for training reward models, which directly uses the data structures discussed.
Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback, Yuntao Bai, Andy Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova DasSarma, Dawn Drain, Stanislav Fort, Deep Ganguli, Tom Henighan, Nicholas Joseph, Saurav Kadavath, Jackson Kernion, Tom Conerly, Sheer El-Showk, Nelson Elhage, Zac Hatfield-Dodds, Danny Hernandez, Tristan Hume, Scott Johnston, Shauna Kravec, Liane Lovitt, Neel Nanda, Catherine Olsson, Dario Amodei, Tom Brown, Jack Clark, Sam McCandlish, Chris Olah, Ben Mann, Jared Kaplan, 2022arXiv preprint arXiv:2204.05862DOI: 10.48550/arXiv.2204.05862 - Introduces the Anthropic HH-RLHF dataset, providing a concrete example of a large-scale preference dataset structured as pairwise comparisons.