Who We Are Looking For
Students and collaborators with backgrounds in machine learning, theory, NLP, vision, HCI, systems, scientific AI, or evaluation.
Join Us
Join the PAI problem.
We are looking for researchers who want to make interpretability provable in theory, actionable in systems, and useful in real workflows.
Opportunities
PhD positions Fall 2027
Postdocs, RAs, visiting PhD students, collaborators Flexible timeline
open_problems.stream()
running Open Problems
What makes an interpretability method provable rather than only plausible?
- 01
What makes an interpretability method provable rather than only plausible?
- 02
How can large-model explanations become actionable evidence for debugging and intervention?
- 03
What systems make interpretability reproducible, comparable, and deployable?
- 04
How can interactive explanations support decisions, control, and human-AI workflows?
- 05
How can PAI methods support scientific hypothesis generation?
submit.application()
How to Apply
Email your CV, academic transcript, and a brief research plan to PAI Lab. Prior publications are not required for PhD applicants; long-term commitment to careful research matters.