Yin Zheng

HOME (Welcome)

Yin Zheng (郑胤)

Researcher, TikTok (ByteDance)

Lead, Code AI Algorithm Team, TikTok AI Innovation Center

Bio

My name is Yin Zheng (郑胤). I am currently a Researcher at TikTok (ByteDance), where I lead the Code AI algorithm team at the TikTok AI Innovation Center. Our mission is to build a general code agent for real-world code performance optimization.

Previously, I led the multilingual large language model (LLM) pretraining initiative at TikTok, building a state-of-the-art multilingual foundation model from scratch for product use—covering end-to-end data pipeline, data mixture strategy, scaling-law driven training, model architecture improvements (including MoE), and a comprehensive multilingual evaluation suite.

Before TikTok, I worked on recommender systems/AutoML/content understanding at Tencent (WeChat Group/AI Lab). I was also a researcher at Hulu focusing on recommender systems.

I received my Ph.D. from Tsinghua University, advised by Prof. Yu-Jin Zhang and Prof. Hugo Larochelle (Mila; previously Google DeepMind/Google Brain).

Research Interests

- Code LLMs & agents for code performance optimization

- Efficient LLM pretraining: data quality, data allocation, scaling laws

- Multilingual foundation models

- Mixture-of-Experts (MoE) and efficient adaptation

- Information Retrieval and Recommender Systems

- AutoML / Neural Architecture Search (NAS)

Contact

Email: yzheng3xg[at]gmail.com

Google Scholar

Selected / Recent Publications

(NeurIPS 2025, poster)

- Exploring Polyglot Harmony: On Multilingual Data Allocation for Large Language Models Pretraining

- MuRating: A High Quality Data Selecting Approach to Multilingual Large Language Model Pretraining

- MoORE: SVD-based Model MoE-ization for Conflict- and Oblivion-Resistant Multi-Task Adaptation

Personal

- I like tennis and I’m a big fan of Roger Federer.

- I like swimming

- Our tennis team won the first prize of tennis championship (马约翰杯) of Tsinghua Univ. in April 2015.

- Vice President of Tencent Tennis Club; organized the Tencent Tennis Championship of 2018.

PROJECTS

1) General Code Agent for Code Performance Optimization (TikTok)

- Lead the Code AI algorithm team at the TikTok AI Innovation Center.

- Mission: build a general code agent that can understand large production codebases and optimize performance with measurable improvements.

2) Multilingual Foundation Model Pretraining (TikTok)

- Built a state-of-the-art multilingual LLM foundation model from scratch for product use.

- End-to-end system: data pipeline, data mixture & allocation strategy, scaling-law guided training, architecture optimization (incl. MoE), multilingual evaluation suite.

3) Recommender Systems / AutoML / Meta Learning(Tencent AI Lab, Hulu)

- Worked on recommender system techniques and AutoML / NAS research.

- Applied deep learning methods to ranking, recommendation, and related applications.

4) Information Retrieval & Content Understanding (Tencent Weixin/WeChat Group)

- Focused on IR and content understanding problems for large-scale applications.

PUBLICATIONS

How to read this page

- “*” indicates equal contribution / co-first author when explicitly noted.

- "**" indicate project lead / Corresponding author

Recent / Highlighted

(NeurIPS 2025, poster)

- Exploring Polyglot Harmony: On Multilingual Data Allocation for Large Language Models Pretraining

Ping Guo, Yubing Ren, Binbin Liu, Fengze Liu, Haobin Lin, Yifan Zhang, Bingni Zhang, Taifeng Wang, Yin Zheng**

- MuRating: A High Quality Data Selecting Approach to Multilingual Large Language Model Pretraining

Zhixun Chen, Ping Guo, Wenhan Han, Yifan Zhang, BINBINLIU, Haobin Lin, Fengze Liu, Yan Zhao, Bingni Zhang, Taifeng Wang, Yin Zheng**, Trevor Cohn, Meng Fang

- MoORE: SVD-based Model MoE-ization for Conflict- and Oblivion-Resistant Multi-Task Adaptation

Shen Yuan, Yin Zheng**, Taifeng Wang, Binbin Liu, Hongteng Xu

Full Publication List

Towards Accurate and Compact Architectures via Neural Architecture Transformer

Yong Guo*, Yin Zheng*, Mingkui Tan*, Qi Chen, Zhipeng Li, Jian Chen, Peilin Zhao, Junzhou Huang

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

A Generic Graph-based Neural Architecture Encoding Scheme for Predictor-based NAS

Xuefei Ning*, Yin Zheng*, Tianchen Zhao, Yu Wang, Huazhong Yang

16TH EUROPEAN CONFERENCE ON COMPUTER VISION (ECCV), 2020

Breaking the Curse of Space Explosion: Towards Efficient NAS with Curriculum Search

Yong Guo*, Yaofo Chen*, Yin Zheng*, Peilin Zhao*, Jian Chen, Junzhou Huang

The International Conference on Machine Learning (ICML), 2020

Towards Fast Adaptation of Neural Architectures with Meta Learning

Dongze Lian*, Yin Zheng*, Yintao Xu, Yanxiong Lu, Leyu Lin, Peilin Zhao, Junzhou Huang, Shenghua Gao

The International Conference on Learning Representations (ICLR), 2020

NAT: Neural Architecture Transformer for Accurate and Compact Architectures

Yong Guo*, Yin Zheng*, Mingkui Tan, Qi Chen, Jian Chen, Peilin Zhao, Junzhou Huang

Advances in Neural Information Processing Systems (NeurIPS), Vancouver, Canada, 2019

50 FPS Object-Level Saliency Detection via Maximally Stable Region

Xiaoming Huang, Yin Zheng, Junzhou Huang, Yu-Jin Zhang

IEEE Transactions on Image Processing

RaFM: Rank-Aware Factorization Machines

Xiaoshuang Chen, Yin Zheng **, Jiaxing Wang, Wenye Ma and Junzhou Huang

The Thirty-sixth International Conference on Machine Learning (ICML), 2019, Long Beach, USA.

A Generalized Locally Linear Factorization Machine with Supervised Variational Encoding

Xiaoshuang Chen, Yin Zheng** ,Peilin Zhao, Zhuxi Jiang, Wenye Ma and Junzhou Huang

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, DOI 10.1109/TKDE.2019.2903403

Fast Single Image Reflection Suppression via Convex Optimization

Yang Yang, Wenye Ma, Yin Zheng, Jian-Feng Cai, Weiyu Xu

IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, USA, 2019

Document Neural Autoregressive Distribution Estimation

Stanislas Lauly, Yin Zheng, Alexandre Allauzen, Hugo Larochelle

Journal of Machine Learning Research, Volume 18, number 113, page 1-24, 2017

Collaborative Filtering with User-Item Co-Autoregressive Models

Chao Du, Chongxuan Li, Yin Zheng, Jun Zhu, Bo Zhang

The Thirty-Second AAAI Conference on Artificial Intelligence (AAAI), 2018

Variational Deep Embedding: An Unsupervised and Generative Approach to Clustering

Zhuxi Jiang, Yin Zheng, Huachun Tan, Bangsheng Tang, Hanning Zhou

International Joint Conference on Artificial Intelligence (IJCAI), Melbourne, Australia, 2017

Neural Autoregressive Collaborative Filtering for Implicit Feedback

Yin Zheng, Cailiang Liu, Bangsheng Tang, Hanning Zhou

Workshop of Deep Learning for Recommender System, in conjunction with the RecSys 2016, Boston.

A Neural Autoregressive Approach to Collaborative Filtering

Yin Zheng, Bangsheng Tang, Wenkui Ding, Hanning Zhou

International Conference on Machine Learning (ICML), 2016

Dynamic Capacity Networks

Amjad Almahairi, Nicolas Ballas, Tim Cooijmans, Yin Zheng, Hugo Larochelle, Aaron Courville

International Conference on Machine Learning (ICML), 2016

A Neural Autoregressive Approach to Attention-based Recognition

Yin Zheng, Richard S. Zemel, Yu-Jin Zhang, Hugo Larochelle

International Journal of Computer Vision (IJCV), 113(1): 67-79, 2015

A Deep and Autoregressive Approach for Topic Modeling of Multimodal Data

Yin Zheng, Yu-Jin Zhang, Hugo Larochelle

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI),

A Locality Preserving Approach for Kernel PCA

Yin Zheng, Bin Shen, Xiaofeng Yang, Wanli Ma, Bao-Di Liu, Yu-Jin Zhang

Proceedings of the 8th International Conference on Image and Graphics (LNCS 9217), 121-135, 2015

Topic Modeling of Multimodal Data: an Autoregressive Approach

Yin Zheng, Yu-Jin Zhang, Hugo Larochelle

IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, Ohio, USA, 2014

A Failure-Aware Explicit Shape Regression Model for Facial Landmark Detection in Video

Meiqing Zhang, Linmi Tao, Yin Zheng, and Yangzhou Du

IEEE SIGNAL PROCESSING LETTERS, VOL. 21, NO. 2, FEBRUARY 2014

Deep Learning and its New Progress in Object and Behavior Recognition

Yin Zheng, Quan-Qi Chen, Yu-Jin Zhang

Journal of Image and Graphics, 2014,19(2):175~184

Discriminant Sparse Coding For Image Classification

Bao-Di Liu, Yu-Xiong Wang, Yu-Jin Zhang, Yin Zheng

IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Kyoto, Japan, 2012.

Action recognition in still images using a combination of human pose and context information

Yin Zheng, Yu-Jin Zhang, Xue Li, Bao-Di Liu

IEEE International Conference on Image Processing (ICIP), Orlando, Florida, USA, 2012