Dissertation Talk: Resolving Information Asymmetry in Multi-Agent Systems

Event details Date: Thursday, July 30, 2026 Time: 9:30 AM to 9:30 AM Location: Gateway 4355 Type: Lecture / Workshop Audience: Faculty,Students About this event Multi-agent systems, from marketplaces to systems of LLM agents, derive much of their richness from information asymmetry: different agents holding different pieces of private information. In this talk, I will draw on theoretical frameworks from economics and machine learning to study how agents’ incentives and interaction dynamics shape the extent to which such gaps can be closed, and how systems might be designed to narrow them. I will show that misaligned incentives can make an information gap impossible to overcome through repeated interaction alone. Even without inherent misalignment, gaps can persist because preferences and rewards are too complex to specify directly. In these settings, carefully designed incentives and learning dynamics can help bridge the gap. Official event details: https://events.berkeley.edu/eecs/event/324868-dissertation-talk-resolving-information-asymmetry