Pith. sign in

REVIEW 9 cited by

Monolith: Real Time Recommendation System With Collisionless Embedding Table

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 2209.07663 v2 pith:MEDKC5QW submitted 2022-09-16 cs.IR

Zhuoran Liu , Leqi Zou , Xuan Zou , Caihua Wang , Biao Zhang , Da Tang , Bolin Zhu , Yijie Zhu
show 3 more authors
Peng Wu Ke Wang Youlong Cheng
This is my paper
classification cs.IR
keywords recommendationsystemframeworksmonolithonlinereal-timecollisionlesscustomer
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Building a scalable and real-time recommendation system is vital for many businesses driven by time-sensitive customer feedback, such as short-videos ranking or online ads. Despite the ubiquitous adoption of production-scale deep learning frameworks like TensorFlow or PyTorch, these general-purpose frameworks fall short of business demands in recommendation scenarios for various reasons: on one hand, tweaking systems based on static parameters and dense computations for recommendation with dynamic and sparse features is detrimental to model quality; on the other hand, such frameworks are designed with batch-training stage and serving stage completely separated, preventing the model from interacting with customer feedback in real-time. These issues led us to reexamine traditional approaches and explore radically different design choices. In this paper, we present Monolith, a system tailored for online training. Our design has been driven by observations of our application workloads and production environment that reflects a marked departure from other recommendations systems. Our contributions are manifold: first, we crafted a collisionless embedding table with optimizations such as expirable embeddings and frequency filtering to reduce its memory footprint; second, we provide an production-ready online training architecture with high fault-tolerance; finally, we proved that system reliability could be traded-off for real-time learning. Monolith has successfully landed in the BytePlus Recommend product.

Discussion (0). Sign in to comment.

Forward citations

Cited by 9 Pith papers

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

  1. Mosaic: A Fleet of User Embedding Specialists for Recommendation at Meta

    cs.IR 2026-07 conditional novelty 6.0 of 10

    Mosaic shows that a fleet of four heterogeneous user-embedding specialists, trained with redundancy-reduction and composite-label losses, improves downstream recommendation quality at Meta.

  2. Multi-Probe Zero Collision Hash (MPZCH): Mitigating Embedding Collisions and Enhancing Model Freshness in Large-Scale Recommenders

    cs.LG 2026-02 unverdicted novelty 6.0 of 10

    MPZCH applies multi-probe linear hashing plus eviction policies to achieve zero collisions on user embeddings and higher freshness on item embeddings while keeping training and inference speeds comparable to standard methods.

  3. xGR: Efficient Generative Recommendation Serving at Scale

    cs.LG 2025-12 conditional novelty 6.0 of 10

    On real-world recommendation datasets, xGR sustains about 2.9–3.5× the throughput of vLLM/xLLM under a 200 ms P99 latency cap through GR-specific KV-cache, beam-search, and scheduling optimizations.

  4. Deep Reinforcement Learning for Ranking Utility Tuning in the Ad Recommender System at Pinterest

    cs.LG 2025-09 conditional novelty 5.0 of 10

    An RL agent that picks personalized weights for a linear ad ranking utility raised treated-segment CTR by 9.7% and CTR30 by 7.7% in Pinterest's production system.

  5. Taming the One-Epoch Phenomenon in Online Recommendation System by Two-stage Contrastive ID Pre-training

    cs.IR 2025-08 conditional novelty 5.0 of 10

    Pre-training ID embeddings with contrastive loss in a simple model avoids one-epoch overfitting and improves Pinterest's recommendation engagement by 2.2%.

  6. LO-FAR: A Cost-Aware Local Filter for Sparse Feature Ranking in Industrial Ad Recommendation

    cs.IR 2026-07 conditional novelty 4.0 of 10

    LO-FAR ranks sparse ID-list features by stand-alone held-out predictive signal and reports downstream NE gains competitive with shuffle importance and BSN at 100–400 retained features in about two CPU-hours.

  7. Mutable Low-Rank Sketches for Retrain-Free Recommendation

    cs.LG 2026-07 conditional novelty 4.0 of 10

    A KP-tree-based mutable sketch lets user embeddings update in O(log n) per rating through a fixed basis, achieving 0.810 RMSE on KuaiRec at 1.8% data read.

  8. A Production-Ready RL Framework for Personalized Utility Tuning with Pareto Sweeping in Pinterest Recommender Systems

    cs.IR 2026-05 unverdicted novelty 4.0 of 10

    PRL-PUTS casts utility-weight tuning as a one-step value-based RL task and uses scalarization-parameter Pareto sweeping at inference time to generate and govern a family of policies, reporting +0.13% lift in successfu...

  9. Joint Model Parameter Scaling and Universal-Domain Data Integration for E-commerce Search Ranking

    cs.IR 2026-03 unverdicted novelty 4.0 of 10

    UniScale couples entire-space data construction with a hierarchical fusion transformer to improve scaling behavior and deliver 1.70% purchase and 2.04% GMV lifts in large-scale e-commerce search A/B tests.

Pith tools