Pith. sign in

REVIEW 8 cited by

MTGR: Industrial-Scale Generative Recommendation Framework in Meituan

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2505.18654 v4 pith:LZX5FAUX submitted 2025-05-24 cs.IR

Ruidong Han , Bin Yin , Shangyu Chen , He Jiang , Fei Jiang , Xiang Li , Chi Ma , Mincong Huang
show 7 more authors
Xiaoguang Li Chunzhen Jing Yueming Han Menglei Zhou Lei Yu Chuan Liu Wei Lin
This is my paper
classification cs.IR
keywords mtgrrecommendationgenerativedlrmfeaturesmeituanmodelscaling
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Scaling law has been extensively validated in many domains such as natural language processing and computer vision. In the recommendation system, recent work has adopted generative recommendations to achieve scalability, but their generative approaches require abandoning the carefully constructed cross features of traditional recommendation models. We found that this approach significantly degrades model performance, and scaling up cannot compensate for it at all. In this paper, we propose MTGR (Meituan Generative Recommendation) to address this issue. MTGR is modeling based on the HSTU architecture and can retain the original deep learning recommendation model (DLRM) features, including cross features. Additionally, MTGR achieves training and inference acceleration through user-level compression to ensure efficient scaling. We also propose Group-Layer Normalization (GLN) to enhance the performance of encoding within different semantic spaces and the dynamic masking strategy to avoid information leakage. We further optimize the training frameworks, enabling support for our models with 10 to 100 times computational complexity compared to the DLRM, without significant cost increases. MTGR achieved 65x FLOPs for single-sample forward inference compared to the DLRM model, resulting in the largest gain in nearly two years both offline and online. This breakthrough was successfully deployed on Meituan, the world's largest food delivery platform, where it has been handling the main traffic.

Discussion (0). Sign in to comment.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. MBGR: Multi-Business Prediction for Generative Recommendation at Meituan

    cs.IR 2026-04 unverdicted novelty 6.0 of 10

    MBGR is a new generative recommendation framework using business-aware semantic IDs, multi-business prediction, and label dynamic routing to handle multiple businesses without seesaw effects or representation confusio...

  2. A Survey on Generative Recommendation: Data, Model, and Tasks

    cs.IR 2025-10 accept novelty 6.0 of 10

    This survey organizes generative recommendation into data, model, and task dimensions, identifying five advantages including world knowledge integration and creative generation while noting challenges in benchmarks an...

  3. Request-Only Optimization for Recommendation Systems

    cs.IR 2025-07 conditional novelty 6.0 of 10

    A request-level training data format eliminates duplicate user features, increasing storage efficiency and training throughput while enabling larger recommendation architectures.

  4. CCFormer: Efficient Cross-Field Interaction and Hierarchical Sequence Compression for Industrial Recommendation at Tencent

    cs.IR 2026-07 conditional novelty 5.0 of 10

    CCFormer beats strong industrial sequential rankers offline and online by pairing field-separated cross-attention with subspace token mixing and hierarchical sequence compression.

  5. OneRec-V2 Technical Report

    cs.IR 2025-08 unverdicted novelty 5.0 of 10

    OneRec-V2 scales generative recommendation to 8B parameters via decoder-only design and real-world preference alignment, improving user engagement metrics in production A/B tests.

  6. An Industrial-Scale Sequential Recommender for LinkedIn Feed Ranking

    cs.IR 2026-02 conditional novelty 4.0 of 10

    A transformer-based sequential recommender, Feed SR, improved LinkedIn Feed time spent by 2.10% in an online A/B test and now serves the majority of Feed traffic.

  7. Coarse-to-Fine Long-term Interest Modeling for Generative Recommendation

    cs.IR 2026-02 conditional novelty 4.0 of 10

    GLASS extends generative retrieval with a tiered long-term interest vector and a first-SID-keyed search of long histories, reporting consistent gains over Tiger and DualGR on two public datasets.

  8. A Survey of Real-World Recommender Systems: Challenges, Constraints, and Industrial Perspectives

    cs.IR 2025-09 conditional novelty 4.0 of 10

    A survey of A/B-validated industrial recommender systems, split into transaction-oriented and content-oriented categories, with a discussion of the academia-industry gap.

Pith tools