Research Engineer, Pre-training Architecture

at  OpenAI

ABOUT THE TEAM

The architecture team is responsible for advancing the neural network architecture of OpenAI’s flagship language models. Our work spans the entire spectrum of architecture development and deployment all the way from the foundational science to the engineering that allows us to train and deploy our models efficiently at scale. Our work has tremendous leverage and deep implications across our stack and we collaborate closely with many other research and engineering teams.

ABOUT THE ROLE

As a Research Engineer, you will develop a deep understanding about model architectures and iterate on research ideas to improve OpenAI’s frontier models. You will systematically study existing architectures and come up with and implement novel ideas to gain understanding and find improvements to our state-of-the-art models.

YOU MIGHT THRIVE IN THIS ROLE IF YOU:

• Are passionate about building mathematical intuitions and performing careful empirical research
• Have strong software engineering skills
• Are familiar with distributed systems and computer architectures
• Enjoy developing deep a understanding of complex systems and collaborating closely with other researchers and engineers