Research

Paper

TESTING March 02, 2026

IDProxy: Cold-Start CTR Prediction for Ads and Recommendation at Xiaohongshu with Multimodal LLMs

Authors

Yubin Zhang, Haiming Xu, Guillaume Salha-Galvan, Ruiyan Han, Feiyang Xiao, Yanhua Huang, Li Lin, Yang Luo, Yao Hu

Abstract

Click-through rate (CTR) models in advertising and recommendation systems rely heavily on item ID embeddings, which struggle in item cold-start settings. We present IDProxy, a solution that leverages multimodal large language models (MLLMs) to generate proxy embeddings from rich content signals, enabling effective CTR prediction for new items without usage data. These proxies are explicitly aligned with the existing ID embedding space and are optimized end-to-end under CTR objectives together with the ranking model, allowing seamless integration into existing large-scale ranking pipelines. Offline experiments and online A/B tests demonstrate the effectiveness of IDProxy, which has been successfully deployed in both Content Feed and Display Ads features of Xiaohongshu's Explore Feed, serving hundreds of millions of users daily.

Metadata

arXiv ID: 2603.01590
Provider: ARXIV
Primary Category: cs.IR
Published: 2026-03-02
Fetched: 2026-03-03 04:34

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.01590v1</id>\n    <title>IDProxy: Cold-Start CTR Prediction for Ads and Recommendation at Xiaohongshu with Multimodal LLMs</title>\n    <updated>2026-03-02T08:16:49Z</updated>\n    <link href='https://arxiv.org/abs/2603.01590v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.01590v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Click-through rate (CTR) models in advertising and recommendation systems rely heavily on item ID embeddings, which struggle in item cold-start settings. We present IDProxy, a solution that leverages multimodal large language models (MLLMs) to generate proxy embeddings from rich content signals, enabling effective CTR prediction for new items without usage data. These proxies are explicitly aligned with the existing ID embedding space and are optimized end-to-end under CTR objectives together with the ranking model, allowing seamless integration into existing large-scale ranking pipelines. Offline experiments and online A/B tests demonstrate the effectiveness of IDProxy, which has been successfully deployed in both Content Feed and Display Ads features of Xiaohongshu's Explore Feed, serving hundreds of millions of users daily.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.IR'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <published>2026-03-02T08:16:49Z</published>\n    <arxiv:primary_category term='cs.IR'/>\n    <author>\n      <name>Yubin Zhang</name>\n    </author>\n    <author>\n      <name>Haiming Xu</name>\n    </author>\n    <author>\n      <name>Guillaume Salha-Galvan</name>\n    </author>\n    <author>\n      <name>Ruiyan Han</name>\n    </author>\n    <author>\n      <name>Feiyang Xiao</name>\n    </author>\n    <author>\n      <name>Yanhua Huang</name>\n    </author>\n    <author>\n      <name>Li Lin</name>\n    </author>\n    <author>\n      <name>Yang Luo</name>\n    </author>\n    <author>\n      <name>Yao Hu</name>\n    </author>\n  </entry>"
}