Auction Design using Value Prediction with Hallucinations

2025-02-12

AI Summary

This paper studies optimal single-item auction design under Bayesian mechanism design with “signal hallucination”: the seller employs a machine learning model to predict bidders’ private values, but the resulting signal may either accurately reflect the true value or be entirely uninformative (i.e., hallucinated). We formally introduce and characterize this hallucination-aware signal model for the first time. Theoretically, for the single-bidder case, the optimal mechanism admits a succinct three-threshold structure—adaptively choosing among “ignore,” “follow,” or “cap” strategies based on signal reliability—yielding strong interpretability and deployability. For multiple bidders, we propose an approximate decomposition framework that decouples the optimal mechanism into two modular components: signal reliability estimation and classical auction design. This decomposition significantly enhances robustness and practicality while preserving near-optimal revenue performance.

Technology Category

• Multiagent Systems: Mechanism Design
• Game Theory and Economic Paradigms: Mechanism Design
• Machine Learning: Online Learning & Bandits

Application Category

• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

We investigate a Bayesian mechanism design problem where a seller seeks to maximize revenue by selling an indivisible good to one of n buyers, incorporating potentially unreliable predictions (signals) of buyers' private values derived from a machine learning model. We propose a framework where these signals are sometimes reflective of buyers' true valuations but other times are hallucinations, which are uncorrelated with the buyers' true valuations. Our main contribution is a characterization of the optimal auction under this framework. Our characterization establishes a near-decomposition of how to treat types above and below the signal. For the one buyer case, the seller's optimal strategy is to post one of three fairly intuitive prices depending on the signal, which we call the"ignore","follow"and"cap"actions.

Problem

• Bayesian mechanism design for revenue maximization
• Incorporating unreliable machine learning predictions
• Optimal auction framework with hallucinations

Innovation

• Bayesian mechanism design
• unreliable value predictions
• optimal auction strategy