معرفی
Caspar Oesterheld is a doctoral student and Researcher at the Computer Science Department of Carnegie Mellon University, supervised by Vincent Conitzer. He focuses on foundational topics in theoretical computer science, game theory, and AI safety. His work includes program equilibrium, decision theory, and cooperative AI.
- Education: Pursuing PhD in Computer Science at CMU
- Advisor: Vincent Conitzer
- Teaching: Co-instructor for CMU CS 15-784 – Foundations of Cooperative AI (Fall 2022), and TA at Duke University (2019–2020).
His research spans game theory, decision theory, and AI safety. Notable contributions include simulation-based program equilibria, analysis of imperfect recall games, and cooperative AI foundations. His work often intersects with multi-agent systems, Pareto improvements, and ethical AI design.
Caspar has published extensively at top venues like AAAI, ICML, and NeurIPS. His recent work explores surrogate goals, observational interference, and dataset creation for decision-theoretic reasoning. He received the AI Alignment Prize and Duke CS Outstanding TA award.
- Grants: FLI AI existential risk fellowship
- Peer-reviewed publications: 15+ papers on game theory, AI safety, and decision theory
- Reviewing: EC, ICML, NeurIPS, and others
