Percy Liang is a Professor of Computer Science at Stanford University, with a courtesy appointment in Statistics. On his faculty page he describes being “drawn to simple things, want to understand things deeply, and like to build useful systems,” with research interests in the fundamental questions around learning and intelligence.
Liang founded the Center for Research on Foundation Models (CRFM) at Stanford, which helped popularize the term “foundation models” for large pre-trained systems like GPT and BERT. His group built HELM (Holistic Evaluation of Language Models), a widely used framework for measuring language models across many tasks and metrics rather than a single score. He also created CodaLab Worksheets, a platform for managing reproducible experiments by tracking their full provenance.
His stated goal is to “enable everyone to understand, shape, and contribute to foundation model development,” reflected in open efforts like the Marin initiative and his popular Stanford course on building language models from scratch. Liang is one of academia’s most prominent advocates for open, rigorous evaluation of frontier AI systems.