The deployment of Large Language Models (LLMs) in recommender systems for predicting Click-Through Rates (CTR) necessitates a delicate balance between computational efficiency and predictive accuracy. This paper presents an optimization framework that combines Retrieval-Augmented Generation (RAG) with an innovative multi-head early exit architecture to concurrently enhance both aspects. By integrating Graph Convolutional Networks (GCNs) as efficient retrieval mechanisms, we are able to significantly reduce data retrieval times while maintaining high model performance. The early exit strategy employed allows for dynamic termination of model inference, utilizing real-time predictive confidence assessments across multiple heads. This not only quickens the responsiveness of LLMs but also upholds or improves their accuracy, making it ideal for real-time application scenarios. Our experiments demonstrate how this architecture effectively decreases computation time without sacrificing the accuracy needed for reliable recommendation delivery, establishing a new standard for efficient, real-time LLM deployment in commercial systems.
@article{arxiv.2501.02173,
title = {The Efficiency vs. Accuracy Trade-off: Optimizing RAG-Enhanced LLM Recommender Systems Using Multi-Head Early Exit},
author = {Huixue Zhou and Hengrui Gu and Xi Liu and Kaixiong Zhou and Mingfu Liang and Yongkang Xiao and Srinivas Govindan and Piyush Chawla and Jiyan Yang and Xiangfei Meng and Huayu Li and Buyun Zhang and Liang Luo and Wen-Yen Chen and Yiping Han and Bo Long and Rui Zhang and Tianlong Chen},
journal= {arXiv preprint arXiv:2501.02173},
year = {2025}
}