Sri Reddy

Sri Reddy

Auto-generated profile, based on publicly available information

Sri Reddy

Meta Platforms, Inc., Menlo Park, CA, USA

Research Focus & Metrics

Representative Publications

View All Publications
13 May 2024
Verified
Conference Paper
Open access
WZWei Zhang
Wei Zhang
DLDai Li
Dai Li
CLChen Liang
Chen Liang
FZFang Zhou
Fang Zhou
ZZZhongke Zhang
Zhongke Zhang
SRSri Reddy
Sri Reddy
Effective user representations are pivotal in personalized advertising. However, stringent constraints on training throughput, serving latency, and memory, often limit the complexity and input feature set of online ads ranking models. This challenge is magnified in extensive systems like Meta's, which encompass hundreds of models with diverse specifications, rendering the tailoring of user representation learning for each model impractical. To address these challenges, we present Scaling User Modeling (SUM), a framework widely deployed in Meta's ads ranking system, designed to facilitate efficient and scalable sharing of online user representation across hundreds of ads models. SUM leverages a few designated upstream user models to synthesize user embeddings from massive amounts of user features with advanced modeling techniques. These embeddings then serve as inputs to downstream online ads ranking models, promoting efficient representation sharing. To adapt to the dynamic nature of user features and ensure embedding freshness, we designed SUM Online Asynchronous Platform (SOAP), a latency-free online serving system complemented with model freshness and embedding stabilization, which enables frequent user model updates and online inference of user embeddings upon each user request. We share our hands-on deployment experiences for the SUM framework and validate its superiority through comprehensive experiments. To date, SUM has been launched to hundreds of ads ranking models in Meta, processing hundreds of billions of user requests daily, yielding significant online metric gains and infrastructure cost savings.

8 October 2024
Verified
Conference Paper
Open access
RXRengan Xu
Rengan Xu
JYJunjie Yang
Junjie Yang
YXYifan Xu
Yifan Xu
HLHong Li
Hong Li
XLXing Liu
Xing Liu
SRSrihari Reddy
Srihari Reddy
The integration of hardware accelerators has significantly advanced the capabilities of modern recommendation systems, enabling the exploration of complex ranking paradigms previously deemed impractical. However, the GPU-based computational costs present substantial challenges. In this paper, we demonstrate our development of an efficiency-driven approach to explore these paradigms, moving beyond traditional reliance on native PyTorch modules. We address the specific challenges posed by ranking models’ dependence on categorical features, which vary in length and complicate GPU utilization. We introduce Jagged Feature Interaction Kernels, a novel method designed to extract fine-grained insights from long categorical features through efficient handling of dynamically sized tensors. We further enhance the performance of attention mechanisms by integrating Jagged tensors with Flash Attention. Our novel Jagged Flash Attention achieves up to 9 × speedup and 22 × memory reduction compared to dense attention. Notably, it also outperforms dense flash attention, with up to 3 × speedup and 53% more memory efficiency. In production models, we observe 10% QPS improvement and 18% memory savings, enabling us to scale our recommendation systems with longer features and more complex architectures.

8 August 2026
Verified
Conference Paper
Open access
LXLee Xiong
Lee Xiong
ZCZhirong Chen
Zhirong Chen
RMRahul Mayuranath
Rahul Mayuranath
SQShangran Qiu
Shangran Qiu
AOAhmet Arda Ozdemir
Ahmet Arda Ozdemir
AOArnold Overwijk
Arnold Overwijk
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 recommendation systems follows predictable power-law scaling similar to LLMs. Crucially, we find that semantic features bend the scaling curve: they are a prerequisite for scaling, enabling the model to effectively utilize the capacity of deeper and longer architectures. To realize the benefits of continued scaling under strict latency constraints, we introduce a two-stage architecture that offloads the heavy computation of large, long-context models to an asynchronous upstream user model. We demonstrate that upstream improvements transfer predictably to downstream ranking tasks. Deployed as the largest user model at Meta, this multi-stage framework drives a 4.3% conversion uplift on Facebook Feed and Reels with minimal serving overhead, establishing a practical blueprint for harnessing scaling laws in industrial recommender systems.

19 July 2026
Verified
Conference Paper
Open access
ZCZikun Cui
Zikun Cui
RWRenzhi Wu
Renzhi Wu
JYJunjie Yang
Junjie Yang
LSLi Sheng
Li Sheng
JWJijie Wei
Jijie Wei
HYHong Yan
Hong Yan

27 September 2026
Verified
Conference Paper
Open access
RWRenzhi Wu
Renzhi Wu
ZCZikun Cui
Zikun Cui
JYJunjie Yang
Junjie Yang
TGTai Guo
Tai Guo
HLHong Li
Hong Li
HYHong Yan
Hong Yan