Training language models to follow instructions with human feedback

AI safety researcher and engineer; former OpenAI employee, founder and AI advisor at Metaculus, and creator of the CosmoTransitions package.

Research scientist at the UK AI Security Institute and former OpenAI member of technical staff who worked on model behavior and post-training research. Previously conducted computational neuroscience research at UC Berkeley.

Sandhini Agarwal is a researcher at OpenAI. Her OpenReview profile lists her as a researcher at OpenAI (2020–present) and an undergraduate student at Stanford University (2015–2019).

Diogo Almeida is an AI researcher and a co-author of the InstructGPT paper (arXiv:2203.02155).

Researcher at OpenAI who led InstructGPT and GPT-4 post-training. He previously co-founded Merlyn Mind and was an engineering director at Quora.

Member of Technical Staff at OpenAI working on machine learning, reinforcement learning, natural language processing, and large language models.

Research scientist at OpenAI working on reinforcement learning and robotics, with a PhD from UC Berkeley.

Engineer who joined OpenAI after building radio encryption systems for small satellites at Planet Labs and working on aerial robots at Airware. He contributed to OpenAI's robotics and reinforcement learning work, including safe exploration research.

Researcher at OpenAI working on language model training and evaluation, and co-author of the GPT-4 Technical Report.

Research scientist at OpenAI focused on multimodal models.

Xu Jiang is a research scientist at OpenAI focused on large reasoning models and multimodal models. Before joining OpenAI, he worked on recommendation systems and search.