👋 My name is Chris Guthrie. I’m currently pursuing a research-based Computer Science Master’s at the University of Colorado in Boulder (my hometown!). I’m also developing a privacy-first RSS recommendation feed called Kitpicks – link coming soon!

CV in Brief

Stanford Undergrad > Travel > App Development Projects (Mobile and Web) > Google > CU Boulder

Research Interests

In general, I’m interested in algorithms which enable computational agents (robots, drones, AI assistants…) to continually learn, adapt, and pursue pro-social goals in “big worlds.”

Theory

• Online learning and optimization
• Bandits and exploration algorithms
• Search/planning algorithms in games and decision processes

One question I’m currently thinking about: When is it suboptimal to model the world as adversarial when it’s actually stochastic/i.i.d.? When do ML practitioners have a good reason not to adopt a “regret minimization” approach?

Practical Techniques

• “Networks of agents” for heterogeneous systems and decentralized decision-making
• LLMs as foundation models, embedding models, and otherwise “components” of decision-making systems
• RL for challenge problems (e.g. Minecraft)