Workshop
AI Agency and Interpretability Workshop
Past event 19–20 September 2026 · Steinberg Auditorium, Baker Hall A36, Adamson Wing, Carnegie Mellon University, Pittsburgh, PA
About
An interdisciplinary meeting on how formal methods from game and decision theory can help interpret, explain, and predict the behavior of frontier AI systems. Sessions cover AI agency and interpretability, belief elicitation, preference elicitation, and instrumental convergence.
Speakers
- Daniel HerrmannPhilosophy, University of North Carolina, Chapel Hill
- Benjamin A. LevinsteinAlignment, Anthropic
- Vincent ConitzerComputer Science, Carnegie Mellon University
- Aran NayebiMachine Learning, Carnegie Mellon University
- Bryan WilderMachine Learning, Carnegie Mellon University
- Aydin MohseniPhilosophy, Carnegie Mellon University
- Sven NethPhilosophy, University of Pittsburgh
Dates
- Saturday, 19 September 2026, 9:00am–5:00pm
- Sunday, 20 September 2026, 9:00am–3:30pm
Steinberg Auditorium, Baker Hall A36, Adamson Wing, Carnegie Mellon University, Pittsburgh, PA.
Attendance was free; registration was requested.
Organizers
Aydin Mohseni, Daniel Herrmann, and Kevin Kelly.
Hosted by the Institute for Complex Social Dynamics, the Center for Formal Epistemology, the Center for Conceptual Research on the Foundations of Safe AI, and the Foundations of Cooperative AI Lab.