Monday, April 20
9:30 am to 11:00 am
Newell-Simon Hall 3305
9:30 am to 11:00 am
Newell-Simon Hall 3305
RI PhD Thesis – Gokul Swamy
Date: April 20, 2026
Time: 9:30-11 AM
Location: NSH 3305
Type: Ph.D. Thesis Defense
Who: Gokul Swamy
Title: Efficient Interactive Learning: Learning More From Less
Abstract: Even as we near the limits of what human-generated data we can scrape from the Internet, today’s decision-making agents — from robots to large language models (LLMs) — are still far from perfect. Thus, as we start to move past the era of simply scaling up training datasets, I believe the most pressing question in decision-making is how we can learn more from less data.
In theory, agents collecting their own data and learning from this experience might allow us to transcend the limits of static datasets. However, there are two core challenges that make delivering on this promise of reinforcement learning (RL) practically challenging. The first is exploration: experiencing the right things. The second is specification: knowing if what you experienced was good or bad. My thesis focuses on algorithms that address both of these challenges in tandem.
This talk will cover both theoretical advancements and their practical implications. First, I will discuss how we can teach robots to provably recover from their own mistakes without extensive trial-and-error exploration. Second, I will describe provably robust algorithms for training language models from conflicting preferences. Third, I will explain how RL learns more from less without having to create data ex nihilo. To conclude, I will outline what I believe are the most promising directions to enable the next generation of agents to learn even more from less.
Committee:
J. Andrew Bagnell (co-chair), Carnegie Mellon University
Zhiwei Steven Wu (co-chair), Carnegie Mellon University
Geoffrey J. Gordon, Carnegie Mellon University
Arthur Gretton, University College London