Efficient LLM-based Advertising via Model Compression and Parallel Verification

2026-05-12

AI Summary

This work addresses the challenge of deploying large language models in advertising systems, where high inference latency and computational costs hinder real-time performance. The authors propose an efficient generative targeted advertising framework that integrates adaptive grouped quantization, layer-wise adaptive hierarchical sparsification, and a trie-based parallel verification mechanism. This novel combination significantly accelerates inference while preserving ad generation quality. Notably, it is the first approach to jointly optimize quantization and sparsification strategies specifically for advertising applications. Evaluated on two real-world datasets, the method achieves substantial speedups—up to ×—with only minimal degradation in output quality, demonstrating strong practical viability for deployment in production advertising systems.

Abstract

Large language models (LLMs) have shown remarkable potential in advertising scenarios such as ad creative generation and targeted advertising. However, deploying LLMs in real-time advertising systems poses significant challenges due to their high inference latency and computational cost. In this paper, we propose an Efficient Generative Targeting framework that integrates adaptive group quantization, layer-adaptive hierarchical sparsification, and prefix-tree parallel verification to accelerate LLM inference while preserving generation quality. Extensive experiments on two real-world advertising scenarios demonstrate that our framework achieves significant speedup with acceptable quality degradation, making it operationally viable for practical deployments.

Problem

• large language models
• advertising
• inference latency
• computational cost
• real-time deployment

Innovation

• adaptive group quantization
• layer-adaptive hierarchical sparsification
• prefix-tree parallel verification
• efficient LLM inference
• generative targeting