Michael J. Curry (UIC)- Truthful Aggregation of LLMs with an Application to Online Advertising
Abstract: The next frontier of online advertising is revenue generation from LLM-generated content. We consider a setting where advertisers aim to influence the responses of an LLM to align with their interests, while platforms seek to maximize advertiser value and ensure user satisfaction. We introduce an auction mechanism over the full output of an LLM that ensures that truthful reporting is a dominant strategy for advertisers. Importantly, the mechanism operates without LLM fine-tuning or access to model weights and provably converges to the output of the optimally fine-tuned LLM as computational resources increase. Additionally, it can incorporate contextual information about advertisers, which significantly improves social welfare. Through experiments with a publicly available LLM, we show that our mechanism leads to high advertiser value and platform revenue with low computational overhead. While our motivating application is online advertising, our mechanism can be applied in any setting with monetary transfers, making it a general-purpose solution for truthfully aggregating the preferences of self-interested agents over LLM-generated replies.
Speakers
Michael J. Curry
Michael J. Curry is an assistant professor of computer science at the University of Illinois Chicago. Prior to this, he was a postdoc at the University of Zurich with Sven Seuken and at Harvard SEAS with David Parkes. He earned his PhD at the University of Maryland, where he was advised by John Dickerson and Tom Goldstein.
Prof. Curry’s research primarily focuses on the intersection of artificial intelligence and strategic incentives. His work has explored the use of machine learning for economic mechanism design (the central theme of his PhD thesis), adversarial robustness in deep learning, the analysis of optimal market-making strategies, and the integration of large language models into mechanism design, among other areas.
