paper-with-me

Papers

Kunlun: Establishing Scaling Laws for Massive-Scale Recommendation Systems through Unified Architecture Design

2026-02-10 · Bojian Hou, Xiaolong Liu, Xiaoyi Liu, Jiaqi Xu, Yasmine Badr, Mengyue Hang, Sudhanshu Chanpuriya, Junqing Zhou, Yuhang Yang, Han Xu, Qiuling Suo, Laming Chen, Yuxi Hu, Jiasheng Zhang, Huaqing Xiong, Yuzhen Huang, Chao Chen, Yue Dong, Yi Yang, Shuo Chang, Xiaorui Gan, Wenlin Chen, Santanu Kolay, Darren Liu, Jade Nie, Chunzhi Yang, Ellie Wen, Jiyan Yang, Huayu Li arxiv

Deriving predictable scaling laws that govern the relationship between model performance and computational investment is crucial for designing and allocating resources in massive-scale recommendation systems. While such laws are established for large language models, they remain challenging for recommendation systems, especially those processing both user history and context features. We identify poor scaling efficiency as the main barrier to predictable power-law scaling, stemming from inefficient modules with low Model FLOPs Utilization (MFU) and suboptimal resource allocation. We introduce Kunlun, a scalable architecture that systematically improves model efficiency and resource allocation. Our low-level optimizations include Generalized Dot-Product Attention (GDPA), Hierarchical Seed Pooling (HSP), and Sliding Window Attention. Our high-level innovations feature Computation Skip (CompSkip) and Event-level Personalization. These advances increase MFU from 17% to 37% on NVIDIA B200 GPUs and double scaling efficiency over state-of-the-art methods. Kunlun is now deployed in major Meta Ads models, delivering significant production impact.

📄 PDF Abstract BibTeX arXiv:2602.10016

Code (0)

등록된 구현이 없습니다.

Tasks

Recommendation Systems

Similar Papers 제목 키워드 기반

Establishing Task Scaling Laws via Compute-Efficient Model Ladders

2024-12-05 · Akshita Bhagia, Jiacheng Liu, Alexander Wettig, David Heineman 외

We develop task scaling laws and model ladders to predict the individual task performance of pretrained language models (LMs) in the overtrained setting. Standard power laws for language modeling loss cannot accurately m…

Language ModelingLanguage ModellingMultiple-choice

Communication-Efficient Language Model Training Scales Reliably and Robustly: Scaling Laws for DiLoCo

2025-03-12 · Zachary Charles, Gabriel Teston, Lucio Dery, Keith Rush 외

As we scale to more massive machine learning models, the frequent synchronization demands inherent in data-parallel approaches create significant slowdowns, posing a critical challenge to further scaling. Recent work dev…

Language ModelingLanguage Modelling

LLaTTE: Scaling Laws for Multi-Stage Sequence Modeling in Large-Scale Ads Recommendation

2026-01-27 · Lee Xiong, Zhirong Chen, Rahul Mayuranath, Shangran Qiu 외 arxiv

We present LLaTTE (LLM-Style Latent Transformers for Temporal Events), a scalable transformer architecture for production ads recommendation. Through systematic experiments, we demonstrate that sequence modeling in recom…

Recommendation Systems

Why Has Predicting Downstream Capabilities of Frontier AI Models with Scale Remained Elusive?

2024-06-06 · Rylan Schaeffer, Hailey Schoelkopf, Brando Miranda, Gabriel Mukobi 외

Predictable behavior from scaling advanced AI systems is an extremely desirable property. Although a well-established literature exists on how pretraining performance scales, the literature on how particular downstream c…

Multiple-choiceQuestion Answering

Fast and Simplex: 2-Simplicial Attention in Triton

2025-07-03 · Aurko Roy, Timothy Chou, Sai Surya Duvvuri, Sijia Chen 외

Recent work has shown that training loss scales as a power law with both model size and the number of tokens, and that achieving compute-optimal models requires scaling model size and token count together. However, these…

valid