# Request-Only Optimization for Recommendation Systems arXiv: https://arxiv.org/abs/2508.05640 ยท HTML: https://mlsanity.com/paper/2508.05640/ ## Key facts - Cited by 5 papers in the map (data through 14 Oct 2026). - Builds on 29 of its 54 references that are in the map. - First posted on arXiv 24 Jul 2025. ## Builds on (most cited first, 29 of 29) - [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017) - [MLP-Mixer: An all-MLP Architecture for Vision](https://mlsanity.com/paper/2105.01601/) (2021) - [Self-Attentive Sequential Recommendation](https://mlsanity.com/paper/1808.09781/) (2018) - [In-Datacenter Performance Analysis of a Tensor Processing Unit](https://mlsanity.com/paper/1704.04760/) (2017) - [Deep Interest Evolution Network for Click-Through Rate Prediction](https://mlsanity.com/paper/1809.03672/) (2018) - [Deep Learning Recommendation Model for Personalization and Recommendation Systems](https://mlsanity.com/paper/1906.00091/) (2019) - [DCN V2: Improved Deep & Cross Network and Practical Lessons for Web-scale Learning to Rank Systems](https://mlsanity.com/paper/2008.13535/) (2020) - [Search-based User Interest Modeling with Lifelong Sequential Behavior Data for Click-Through Rate Prediction](https://mlsanity.com/paper/2006.05639/) (2020) - [Practice on Long Sequential User Behavior Modeling for Click-Through Rate Prediction](https://mlsanity.com/paper/1905.09248/) (2019) - [Multi-Interest Network with Dynamic Routing for Recommendation at Tmall](https://mlsanity.com/paper/1904.08030/) (2019) - [Learning Tree-based Deep Model for Recommender Systems](https://mlsanity.com/paper/1801.02294/) (2018) - [Software-Hardware Co-design for Fast and Scalable Training of Deep Learning Recommendation Models](https://mlsanity.com/paper/2104.05158/) (2021) - [PinnerFormer: Sequence Modeling for User Representation at Pinterest](https://mlsanity.com/paper/2205.04507/) (2022) - [DeepRecSys: A System for Optimizing End-To-End At-scale Neural Recommendation Inference](https://mlsanity.com/paper/2001.02772/) (2020) - [MTGR: Industrial-Scale Generative Recommendation Framework in Meituan](https://mlsanity.com/paper/2505.18654/) (2025) - [LONGER: Scaling Up Long Sequence Modeling in Industrial Recommenders](https://mlsanity.com/paper/2505.04421/) (2025) - [TWIN: TWo-stage Interest Network for Lifelong User Behavior Modeling in CTR Prediction at Kuaishou](https://mlsanity.com/paper/2302.02352/) (2023) - [RecSSD: Near Data Processing for Solid State Drive Based Recommendation Inference](https://mlsanity.com/paper/2102.00075/) (2021) - [Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Training](https://mlsanity.com/paper/2108.09373/) (2021) - [Sampling Is All You Need on Modeling Long-Term User Behaviors for CTR Prediction](https://mlsanity.com/paper/2205.10249/) (2022) - [MLP4Rec: A Pure MLP Architecture for Sequential Recommendations](https://mlsanity.com/paper/2204.11510/) (2022) - [Learning Piece-wise Linear Models from Large Scale Data for Ad Click Prediction](https://mlsanity.com/paper/1704.05194/) (2017) - [Rethinking Personalized Ranking at Pinterest: An End-to-End Approach](https://mlsanity.com/paper/2209.08435/) (2022) - [Serving Deep Learning Models with Deduplication from Relational Databases](https://mlsanity.com/paper/2201.10442/) (2022) - [Revisiting Neural Retrieval on Accelerators](https://mlsanity.com/paper/2306.04039/) (2023) - [RecD: Deduplication for End-to-End Deep Learning Recommendation Model Training Infrastructure](https://mlsanity.com/paper/2211.05239/) (2022) - [Scaling Recommender Transformers to One Billion Parameters](https://mlsanity.com/paper/2507.15994/) (2025) - [Context-based Fast Recommendation Strategy for Long User Behavior Sequence in Meituan Waimai](https://mlsanity.com/paper/2403.12566/) (2024) - [Retrieval with Learned Similarities](https://mlsanity.com/paper/2407.15462/) (2024) ## Built on by (most cited first, 5 of 5) - [Target-Aware Early Stage Ranking](https://mlsanity.com/paper/2511.21095/) (2025) - [Versioned Late Materialization for Ultra-Long Sequence Training in Recommendation Systems at Scale](https://mlsanity.com/paper/2604.24806/) (2026) - [Context Features Are Cheap: Rank-Aware Decomposition for Efficient Feature Interaction in Recommender Systems](https://mlsanity.com/paper/2605.27450/) (2026) - [WHALE: A Scalable Unified Model for Recommendation with Wukong-HSTU Architecture](https://mlsanity.com/paper/2607.17017/) (2026) - [ROCS: Request-Oriented Compute Sharing for Efficient Large-Scale Recommendation](https://mlsanity.com/paper/2607.27744/) (2026) ## Abstract (arXiv) Deep Learning Recommendation Models (DLRMs) represent one of the largest machine learning applications on the planet. Industry-scale DLRMs are trained with petabytes of recommendation data to serve billions of users every day. To utilize the rich user signals in the long user history, DLRMs have been scaled up to unprecedented complexity, up to trillions of floating-point operations (TFLOPs) per example. This scale, coupled with the huge amount of training data, necessitates new storage and training algorithms to efficiently improve the quality of these complex recommendation systems. In this paper, we present a Request-Only Optimizations (ROO) training and modeling paradigm. ROO simultaneously improves the storage and training efficiency as well as the model quality of recommendation systems. We holistically approach this challenge through co-designing data (i.e., request-only data), infrastructure (i.e., request-only based data processing pipeline), and model architecture (i.e., request-only neural architectures). Our ROO training and modeling paradigm treats a user request as a unit of the training data. Compared with the established practice of treating a user impression as a unit, our new design achieves native feature deduplication in data logging, consequently saving data storage. Second, by de-duplicating computations and communications across multiple impressions in a request, this new paradigm enables highly scaled-up neural network architectures to better capture user interest signals, such as Generative Recommenders (GRs) and other request-only friendly architectures.