Wen-Yen Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Wen-Yen Chen has a total of 68 co-authors. Here are the top co-authors by number of shared publications:
- Wei Wen
Meta AI
4 shared publications - Xi Liu
Meta Platforms, Sunnyvale, CA, USA
4 shared publications - Buyun Zhang
Meta AI, Menlo Park, USA
3 shared publications - Hang Yin
Meta Platforms, Inc. Menlo Park, USA
3 shared publications - Huayu Li
Meta Platforms, Inc., Menlo Park, CA, USA
3 shared publications - Jiyan Yang
Meta Platforms, Inc., Menlo Park, CA, USA
3 shared publications
Latest Publications
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4000-4011
Generalizable Multi-Pass Training of Ads Recommendation Models with Foundation Model Guidance
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4000-4011
Open AccessGeneralizable Multi-Pass Training of Ads Recommendation Models with Foundation Model Guidance
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 6469-6480
Mixture of Sequence: Theme-Aware Mixture-of-Experts for Long-Sequence Recommendation
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 6469-6480
Open AccessMixture of Sequence: Theme-Aware Mixture-of-Experts for Long-Sequence Recommendation
- Top Cited PapersOpen Access
- 10 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 6225-6233
InterFormer: Effective Heterogeneous Interaction Learning for Click-Through Rate Prediction
- 10 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 6225-6233
Top Cited PapersOpen AccessInterFormer: Effective Heterogeneous Interaction Learning for Click-Through Rate Prediction
- Open Access
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4536-4547
SiGeo: Sub-One-Shot NAS via Geometry of Loss Landscape
Neural Architecture Search (NAS) has become a widely used tool for automating neural network design. While one-shot NAS methods have successfully reduced computational requirements, they often require extensive training. On the other hand, zero-shot NAS utilizes training-free proxies to evaluate a candidate architecture's test performance but has two limitations: (1) inability to use the information gained as a network improves with training and (2) unreliable performance, particularly in complex domains like RecSys, due to the multi-modal data inputs and complex architecture configurations. To synthesize the benefits of both methods, we introduce a "sub-one-shot" paradigm that serves as a bridge between zero-shot and one-shot NAS. In sub-one-shot NAS, the supernet is trained using only a small subset of the training data, a phase we refer to as "warm-up." Within this framework, we present SiGeo, a proxy founded on a novel theoretical framework that connects the supernet warm-up with the efficacy of the proxy. Extensive experiments have consistently shown that SiGeo, when properly warmed up, surpasses state-of-the-art NAS proxies in many established NAS benchmarks in the computer vision domain. Furthermore, when tested on recommendation system benchmarks, SiGeo demonstrates its ability to match the performance of state-of-the-art weight-sharing one-shot NAS methods while significantly reducing computational costs by approximately 60%.
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4536-4547
Open AccessSiGeo: Sub-One-Shot NAS via Geometry of Loss Landscape
Neural Architecture Search (NAS) has become a widely used tool for automating neural network design. While one-shot NAS methods have successfully reduced computational requirements, they often require extensive training. On the other hand, zero-shot NAS utilizes training-free proxies to evaluate a candidate architecture's test performance but has two limitations: (1) inability to use the information gained as a network improves with training and (2) unreliable performance, particularly in complex domains like RecSys, due to the multi-modal data inputs and complex architecture configurations. To synthesize the benefits of both methods, we introduce a "sub-one-shot" paradigm that serves as a bridge between zero-shot and one-shot NAS. In sub-one-shot NAS, the supernet is trained using only a small subset of the training data, a phase we refer to as "warm-up." Within this framework, we present SiGeo, a proxy founded on a novel theoretical framework that connects the supernet warm-up with the efficacy of the proxy. Extensive experiments have consistently shown that SiGeo, when properly warmed up, surpasses state-of-the-art NAS proxies in many established NAS benchmarks in the computer vision domain. Furthermore, when tested on recommendation system benchmarks, SiGeo demonstrates its ability to match the performance of state-of-the-art weight-sharing one-shot NAS methods while significantly reducing computational costs by approximately 60%.
- Open Access
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 73-82
Rankitect: Ranking Architecture Search Battling World-class Engineers at Meta Scale
Neural Architecture Search (NAS) has demonstrated its efficacy in computer vision and potential for ranking systems. However, prior work focused on academic problems, which are evaluated at small scale under well-controlled fixed baselines. In industry system, such as ranking system in Meta, it is unclear whether NAS algorithms from the literature can outperform production baselines because of: (1) scale - Meta's ranking systems serve billions of users, (2) strong baselines - the baselines are production models optimized by hundreds to thousands of world-class engineers for years since the rise of deep learning, (3) dynamic baselines - engineers may have established new and stronger baselines during NAS search, and (4) efficiency - the search pipeline must yield results quickly in alignment with the productionization life cycle. In this paper, we present Rankitect, a NAS software framework for ranking systems at Meta. Rankitect seeks to build brand new architectures by composing low level building blocks from scratch. Rankitect implements and improves state-of-the-art (SOTA) NAS methods for comprehensive and fair comparison under the same search space, including sampling-based NAS, one-shot NAS, and Differentiable NAS (DNAS). We evaluate Rankitect by comparing to multiple production ranking models at Meta. We find that Rankitect can discover new models from scratch achieving competitive trade-off between Normalized Entropy loss and FLOPs. When utilizing search space designed by engineers, Rankitect can generate better models than engineers, achieving positive offline evaluation and online A/B test at Meta scale.
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 73-82
Open AccessRankitect: Ranking Architecture Search Battling World-class Engineers at Meta Scale
Neural Architecture Search (NAS) has demonstrated its efficacy in computer vision and potential for ranking systems. However, prior work focused on academic problems, which are evaluated at small scale under well-controlled fixed baselines. In industry system, such as ranking system in Meta, it is unclear whether NAS algorithms from the literature can outperform production baselines because of: (1) scale - Meta's ranking systems serve billions of users, (2) strong baselines - the baselines are production models optimized by hundreds to thousands of world-class engineers for years since the rise of deep learning, (3) dynamic baselines - engineers may have established new and stronger baselines during NAS search, and (4) efficiency - the search pipeline must yield results quickly in alignment with the productionization life cycle. In this paper, we present Rankitect, a NAS software framework for ranking systems at Meta. Rankitect seeks to build brand new architectures by composing low level building blocks from scratch. Rankitect implements and improves state-of-the-art (SOTA) NAS methods for comprehensive and fair comparison under the same search space, including sampling-based NAS, one-shot NAS, and Differentiable NAS (DNAS). We evaluate Rankitect by comparing to multiple production ranking models at Meta. We find that Rankitect can discover new models from scratch achieving competitive trade-off between Normalized Entropy loss and FLOPs. When utilizing search space designed by engineers, Rankitect can generate better models than engineers, achieving positive offline evaluation and online A/B test at Meta scale.
- Open Access
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 374-382
AutoML for Large Capacity Modeling of Meta's Ranking Systems
Web-scale ranking systems at Meta serving billions of users is complex. Improving ranking models is essential but engineering heavy. Automated Machine Learning (AutoML) can potentially release engineers from labor intensive work of tuning ranking models; however, it is unknown if AutoML is efficient enough to meet tight production timeline in real-world applications and, at the same time, bring additional improvements to the already strong baselines. Moreover, to achieve higher ranking performance, there is an ever-increasing demand to scale up ranking models to even larger capacity, which imposes more challenges on the AutoML efficiency. The large scale of models and tight production schedule requires AutoML to outperform human baselines by only using a small number of model evaluation trials (~100). This paper presents a sampling-based AutoML search method, focusing on neural architecture search and hyperparameter optimization, with a particular emphasis on addressing aforementioned challenges in Meta-scale production when building large capacity models. Our approach efficiently handles large-scale data demands. It leverages a lightweight predictor-based searcher and reinforcement learning to explore vast search spaces, significantly reducing the number of model evaluations. Through experiments in large capacity modeling for CTR and CVR applications, we have demonstrated that our method achieves outstanding Return-on-Investment (ROI) versus human tuned baselines, with up to 0.09% Normalized Entropy (NE) loss reduction or 25% Query-per-Second (QPS) increase by only sampling one hundred models on average from a curated search space. The proposed AutoML method has already made real-world impact where a discovered Instagram CTR model with up to -0.36% NE gain (over existing production baseline) was selected for large-scale online A/B test and show statistically significant gain. These production results proved AutoML efficacy and accelerated its adoption in ranking systems at Meta.
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 374-382
Open AccessAutoML for Large Capacity Modeling of Meta's Ranking Systems
Web-scale ranking systems at Meta serving billions of users is complex. Improving ranking models is essential but engineering heavy. Automated Machine Learning (AutoML) can potentially release engineers from labor intensive work of tuning ranking models; however, it is unknown if AutoML is efficient enough to meet tight production timeline in real-world applications and, at the same time, bring additional improvements to the already strong baselines. Moreover, to achieve higher ranking performance, there is an ever-increasing demand to scale up ranking models to even larger capacity, which imposes more challenges on the AutoML efficiency. The large scale of models and tight production schedule requires AutoML to outperform human baselines by only using a small number of model evaluation trials (~100). This paper presents a sampling-based AutoML search method, focusing on neural architecture search and hyperparameter optimization, with a particular emphasis on addressing aforementioned challenges in Meta-scale production when building large capacity models. Our approach efficiently handles large-scale data demands. It leverages a lightweight predictor-based searcher and reinforcement learning to explore vast search spaces, significantly reducing the number of model evaluations. Through experiments in large capacity modeling for CTR and CVR applications, we have demonstrated that our method achieves outstanding Return-on-Investment (ROI) versus human tuned baselines, with up to 0.09% Normalized Entropy (NE) loss reduction or 25% Query-per-Second (QPS) increase by only sampling one hundred models on average from a curated search space. The proposed AutoML method has already made real-world impact where a discovered Instagram CTR model with up to -0.36% NE gain (over existing production baseline) was selected for large-scale online A/B test and show statistically significant gain. These production results proved AutoML efficacy and accelerated its adoption in ranking systems at Meta.