Saved in:
Bibliographic Details
Main Authors: Dong, Wenxin, Gao, Chang, Yu, Guanghui, Jiao, Xuewu, Hu, Mingqing, Fu, Qiang, Xu, Peng, Wei, Penghui, Xu, Hui, Xing, Yue, Li, Shuanglong, Liu, Lin
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.11582
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910211197894656
author Dong, Wenxin
Gao, Chang
Yu, Guanghui
Jiao, Xuewu
Hu, Mingqing
Fu, Qiang
Xu, Peng
Wei, Penghui
Xu, Hui
Xing, Yue
Li, Shuanglong
Liu, Lin
author_facet Dong, Wenxin
Gao, Chang
Yu, Guanghui
Jiao, Xuewu
Hu, Mingqing
Fu, Qiang
Xu, Peng
Wei, Penghui
Xu, Hui
Xing, Yue
Li, Shuanglong
Liu, Lin
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2605_11582
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Efficient LLM-based Advertising via Model Compression and Parallel Verification
Dong, Wenxin
Gao, Chang
Yu, Guanghui
Jiao, Xuewu
Hu, Mingqing
Fu, Qiang
Xu, Peng
Wei, Penghui
Xu, Hui
Xing, Yue
Li, Shuanglong
Liu, Lin
Computation and Language
I.2.7; H.3.5
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.
title Efficient LLM-based Advertising via Model Compression and Parallel Verification
topic Computation and Language
I.2.7; H.3.5
url https://arxiv.org/abs/2605.11582