Yin, Yutong (尹禹童)

PhD Student, Department of Industrial Engineering and Management Sciences, Northwestern University, Evanston, IL, USA E-mail: yutongyin2028@u.northwestern.edu Phone: +1 773-250-4677

About Me

I am a third-year PhD student at Northwestern University under the supervision of Professor Zhaoran Wang. Prior to this, I earned my bachelor's degree in Artificial Intelligence from Peking University. During my undergraduate studies, I served as a research assistant to Professor Xiaotie Deng at Peking University and collaborated with Dr. Zhijian Duan. In my senior year, I had the privilege of interning with Professor Zhuoran Yang at Yale University and Professor Zhaoran Wang at Northwestern University.

My research focuses on machine learning theory, deep learning, large language models, and mechanistic interpretability. I am particularly dedicated to unraveling the mechanisms that contribute to the exceptional performance of machine learning models and algorithms, and leveraging this understanding to enhance their effectiveness.

Education

• Northwestern University, Evanston, IL, USA
  PhD in Industrial Engineering and Management Sciences | Sept 2023 – Now
  
• Peking University, Beijing, China
  B.E. in Artificial Intelligence, Yuanpei College | Sept 2019 – July 2023
  

Publications

1. Zuo, Y., Yin, Y., Zeng, Z., Li, A., Zhu, B., & Wang, Z. “Local Linear Attention: An Optimal Interpolation of Linear and Softmax Attention For Test-Time Regression.”. In Proceedings of the International Conference on Learning Representations (ICLR), 2026. [arXiv]
2. Yin, Yutong, and Zhaoran Wang. “Are Transformers Able to Reason by Connecting Separated Knowledge in Training Data?”. In Proceedings of the International Conference on Learning Representations (ICLR), 2025. [arXiv / Code]
3. Zhong, H., Yin, Y., Zhang, S., et al. “BRiTE: Bootstrapping Reinforced Thinking Process to Enhance Language Model Reasoning.” In Proceedings of the 42nd International Conference on Machine Learning (ICML), 2025. [arXiv]
4. Duan, Z., Tang, J., Yin, Y., Feng, Z., Yan, X., Zaheer, M., & Deng, X. “A Context-Integrated Transformer-Based Neural Network for Auction Design.” In Proceedings of the 39th International Conference on Machine Learning (ICML), 2022. [arXiv / Code]

Work Experience

• ByteDance, San Jose, USA
  Research Scientist Intern, AI Lab | Jan 2025 – Mar 2025
  • Researched LLM mechanistic interpretability (token-wise learning dynamics, latent concept formation).
  • Built multi-GPU/multi-node pretraining/fine-tuning pipelines for rapid ablations and reproducible results.
  • Applied Sparse Autoencoders to probe latent features.

Skills

• Programming: C++, Python, LaTeX
• Frameworks: PyTorch, CUDA, Git
• Advanced Knowledge: Statistics and Opmization Theory, Machine Learning Theory, LLM Fine-tuning, Deep Learning Interpretability, Reinforcement Learning
• Languages: GRE 324, TOEFL 107