Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.16474 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916017804935168 |
|---|---|
| author | Sun, Haoran Song, Xinrui Zhang, Xinyu Chen, Zhaohua Chu, Xu Zhang, Zhilin Yu, Chuan Xu, Jian Zheng, Bo Deng, Xiaotie |
| author_facet | Sun, Haoran Song, Xinrui Zhang, Xinyu Chen, Zhaohua Chu, Xu Zhang, Zhilin Yu, Chuan Xu, Jian Zheng, Bo Deng, Xiaotie |
| contents | The integration of advertising auction mechanisms into large language model (LLM)-based chatbots presents a significant opportunity for commercialization, yet poses unique challenges in balancing relevance, efficiency, and user experience. Recently, Feizi et al.~\citep{feizi2023online} and Hajiaghayi et al.~\citep{hajiaghayi2024ad} outlined a retrieve-then-generate paradigm that decouples retrieval and generation, offering lightweight ad insertion and payment determination. However, current retrieval relies solely on text embedding similarity, which may lead to commercial misinterpretation and issues such as repetitive insertions. In this paper, we propose LERA, a two-stage retrieve-then-generate auction framework tailored for LLM chatbots. In the first stage, embedding-based coarse filtering pre-selects a small set of candidate advertisers. In the second stage, the LLM itself is queried with a carefully designed prompt to produce logits over candidates, which serve as refined organic relevance scores. These scores are combined with bids, and a critical-value payment rule accounts for both the coarse-filtering and fine-ranking thresholds, ensuring truthfulness for utility-maximizing advertisers. The framework naturally extends to multiple ad insertions within dynamic dialogue flows and long responses. Experiments on a synthetic advertiser-query benchmark show that LERA substantially improves ad selection accuracy and insertion diversity while incurring only controllable latency overhead. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_16474 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | LERA: LLM-Enhanced RAG for Ad Auction in Generative Chatbots Sun, Haoran Song, Xinrui Zhang, Xinyu Chen, Zhaohua Chu, Xu Zhang, Zhilin Yu, Chuan Xu, Jian Zheng, Bo Deng, Xiaotie Information Retrieval Artificial Intelligence The integration of advertising auction mechanisms into large language model (LLM)-based chatbots presents a significant opportunity for commercialization, yet poses unique challenges in balancing relevance, efficiency, and user experience. Recently, Feizi et al.~\citep{feizi2023online} and Hajiaghayi et al.~\citep{hajiaghayi2024ad} outlined a retrieve-then-generate paradigm that decouples retrieval and generation, offering lightweight ad insertion and payment determination. However, current retrieval relies solely on text embedding similarity, which may lead to commercial misinterpretation and issues such as repetitive insertions. In this paper, we propose LERA, a two-stage retrieve-then-generate auction framework tailored for LLM chatbots. In the first stage, embedding-based coarse filtering pre-selects a small set of candidate advertisers. In the second stage, the LLM itself is queried with a carefully designed prompt to produce logits over candidates, which serve as refined organic relevance scores. These scores are combined with bids, and a critical-value payment rule accounts for both the coarse-filtering and fine-ranking thresholds, ensuring truthfulness for utility-maximizing advertisers. The framework naturally extends to multiple ad insertions within dynamic dialogue flows and long responses. Experiments on a synthetic advertiser-query benchmark show that LERA substantially improves ad selection accuracy and insertion diversity while incurring only controllable latency overhead. |
| title | LERA: LLM-Enhanced RAG for Ad Auction in Generative Chatbots |
| topic | Information Retrieval Artificial Intelligence |
| url | https://arxiv.org/abs/2605.16474 |