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

TL;DR

IDProxy utilizes multimodal large language models to generate proxy embeddings from rich content signals, effectively addressing cold-start challenges in CTR prediction for ads and recommendations at Xiaohongshu.

Contribution

This work introduces IDProxy, a novel approach that aligns multimodal LLM-generated proxies with existing ID embeddings for improved cold-start CTR prediction.

Findings

Effective in cold-start scenarios without usage data

Successfully deployed in Xiaohongshu's large-scale systems

Demonstrated improvements in offline and online experiments

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.