用稀疏化压缩推荐系统嵌入，省内存还保效果。
The Future is Sparse: Embedding Compression for Scalable Retrieval in Recommender Systems
- 将稠密嵌入映射到高维稀疏激活空间，实现轻量压缩。
- 在严格资源限制下仍保持检索性能，支持大规模部署。
- 适合追求高效可扩展推荐系统的工程团队使用。
工业级推荐系统面临核心挑战：如何用密集嵌入表示高基数实体（如用户或物品），且需在训练和推理时均可访问。但随着嵌入规模增大，内存限制使存储与访问日益困难。本文提出一种轻量级、可学习的嵌入压缩技术，将稠密嵌入投影至高维稀疏激活空间。该方法专为检索任务设计，在降低内存需求的同时保持检索性能，使系统在严苛资源约束下仍可规模化部署。实验表明，利用稀疏性是提升大规模推荐系统效率的可行路径。代码已开源：https://github.com/recombee/CompresSAE。
原文摘要 · Abstract (English)
Industry-scale recommender systems face a core challenge: representing entities with high cardinality, such as users or items, using dense embeddings that must be accessible during both training and inference. However, as embedding sizes grow, memory constraints make storage and access increasingly difficult. We describe a lightweight, learnable embedding compression technique that projects dense embeddings into a high-dimensional, sparsely activated space. Designed for retrieval tasks, our method reduces memory requirements while preserving retrieval performance, enabling scalable deployment under strict resource constraints. Our results demonstrate that leveraging sparsity is a promising approach for improving the efficiency of large-scale recommenders. We release our code at https://github.com/recombee/CompresSAE.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。