I do research primarily on Large Language Models (LLMs), ranging from topics such as efficient at-scale distributed training, to critical evaluation work of AI systems, to steering and interpreting AI/ML systems' inner workings. I am a contributor or maintainer to several open-source widely used libraries, including GPT-NeoX and the LM Evaluation Harness.

At EleutherAI, my job includes: publishing research on NLP + LLMs in top academic venues as part of a non-profit research lab ; open-source code contributions ; and managing research projects with volunteer and full-time employee contributors.