Ad Auctions for LLMs via Retrieval Augmented Generation

2024-06-12

Neural Information Processing Systems

AI Summary

This paper addresses the tension between ad embedding and revenue optimization in large language model (LLM) text generation. Methodologically, it introduces a paragraph-level ad auction mechanism that integrates retrieval-augmented generation (RAG) with game-theoretic auction design. It formally defines a log-social-welfare maximization objective—uniquely balancing allocative efficiency and fairness—and constructs an incentive-compatible framework for joint multi-ad allocation and dynamic pricing, supporting probabilistic ad retrieval and paragraph-level relevance modeling. Theoretically, the mechanism is proven to satisfy incentive compatibility and achieve log-social-welfare optimality. Empirically, it significantly improves the trade-off between advertising revenue and content relevance across diverse scenarios, uncovering critical insights into flexibility–performance trade-offs and metric balancing.

Technology Category

• Search and Optimization: Learning to Search
• Multiagent Systems: Mechanism Design
• Game Theory and Economic Paradigms: Mechanism Design

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: Ad search and search for Web retail
• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising

Abstract

In the field of computational advertising, the integration of ads into the outputs of large language models (LLMs) presents an opportunity to support these services without compromising content integrity. This paper introduces novel auction mechanisms for ad allocation and pricing within the textual outputs of LLMs, leveraging retrieval-augmented generation (RAG). We propose a segment auction where an ad is probabilistically retrieved for each discourse segment (paragraph, section, or entire output) according to its bid and relevance, following the RAG framework, and priced according to competing bids. We show that our auction maximizes logarithmic social welfare, a new notion of welfare that balances allocation efficiency and fairness, and we characterize the associated incentive-compatible pricing rule. These results are extended to multi-ad allocation per segment. An empirical evaluation validates the feasibility and effectiveness of our approach over several ad auction scenarios, and exhibits inherent tradeoffs in metrics as we allow the LLM more flexibility to allocate ads.

Problem

• Develop auction mechanisms for ad placement in LLM outputs
• Balance ad relevance and bid fairness via retrieval-augmented generation
• Optimize multi-ad allocation per segment while ensuring incentive compatibility

Innovation

• RAG-based ad auction for LLM outputs
• Segment auction with probabilistic ad retrieval
• Logarithmic social welfare maximization