5 sources
One of the clearest people to follow if you care about scaling laws, training efficiency, and the systems choices that quietly shape frontier-model progress.
Researcher Profile
Editor reviewedPercy Liang
Evaluation, robust NLP, systems
Professor at Stanford University
A key person for understanding how foundation-model evaluation, governance, and research tooling became a coherent agenda rather than a scattered set of concerns.
Organizations
Stanford University
About This Page
This profile is meant to help you get oriented quickly: why this researcher matters, what to read first, and where to explore next.
Last reviewed
March 18, 2026
Known For
The ideas, systems, and research directions that make this person worth knowing.
01
Foundation-model evaluation and governance
02
HELM and CRFM
03
Building practical research infrastructure around model assessment
04
Evaluation, robust NLP, systems
05
Percy Liang (site)
06
NLP
Start Here
Canonical papers, project pages, or repositories that anchor this profile.
Signature Works
Additional papers, projects, or repositories that help flesh out the profile.
Supporting Sources
Additional links that help verify and flesh out this profile.
Related Researchers
People worth exploring next because they share topics, labs, or source material with this profile.
4 sources
A useful person to follow for the part of Anthropic that moved from assistant training into explicit behavior-discovery and evaluation work.
4 sources
Important because his work sits near the point where technical alignment, evaluation practice, and the public case for safer frontier-model deployment meet.
4 sources
Worth following for the evaluation side of Anthropic’s alignment program, especially where model-written tests are used to surface new behaviors quickly.
4 sources
A good person to follow if you care about the practical evaluation layer at Anthropic rather than only its highest-level alignment claims.
4 sources
A useful profile for the segment of Anthropic that turns alignment concerns into concrete evals, behavior probes, and red-team-style measurement.