Publications

2025

1. A Theory of Learning with Autoregressive Chain of Thought

   Nirmit Joshi, Gal Vardi, Adam Block, Surbhi Goel, Zhiyuan Li, Theodor Misiakiewicz, and Nathan Srebro

   In Proceedings of the 38th Conference on Learning Theory, COLT 2025

   arXiv Bib

   @inproceedings{joshi2025theory,
     author = {Joshi, Nirmit and Vardi, Gal and Block, Adam and Goel, Surbhi and Li, Zhiyuan and Misiakiewicz, Theodor and Srebro, Nathan},
     title = {A Theory of Learning with Autoregressive Chain of Thought},
     booktitle = {Proceedings of the 38th Conference on Learning Theory},
     year = {2025},
   }

2. Structured Preconditioners in Adaptive Optimization: A Unified Analysis

   Shuo Xie, Tianhao Wang, Sashank Reddi, Sanjiv Kumar, and Zhiyuan Li

   In Proceedings of the 42nd International Conference on Machine Learning, ICML 2025

   arXiv Bib

   @inproceedings{xie2025structured,
     author = {Xie, Shuo and Wang, Tianhao and Reddi, Sashank and Kumar, Sanjiv and Li, Zhiyuan},
     title = {Structured Preconditioners in Adaptive Optimization: A Unified Analysis},
     booktitle = {Proceedings of the 42nd International Conference on Machine Learning},
     year = {2025},
   }

3. PENCIL: Long Thoughts with Short Memory

   Chenxiao Yang, Nathan Srebro, David McAllester, and Zhiyuan Li

   In Proceedings of the 42nd International Conference on Machine Learning, ICML 2025

   arXiv Bib

   @inproceedings{yang2025pencil,
     author = {Yang, Chenxiao and Srebro, Nathan and McAllester, David and Li, Zhiyuan},
     title = {PENCIL: Long Thoughts with Short Memory},
     booktitle = {Proceedings of the 42nd International Conference on Machine Learning},
     year = {2025},
   }

4. Weak-to-Strong Generalization Even in Random Feature Networks, Provably

   Marko Medvedev, Kaifeng Lyu, Dingli Yu, Sanjeev Arora, Zhiyuan Li, and Nathan Srebro

   In Proceedings of the 42nd International Conference on Machine Learning, ICML 2025

   arXiv Bib

   @inproceedings{medvedev2025weak,
     author = {Medvedev, Marko and Lyu, Kaifeng and Yu, Dingli and Arora, Sanjeev and Li, Zhiyuan and Srebro, Nathan},
     title = {Weak-to-Strong Generalization Even in Random Feature Networks, Provably},
     booktitle = {Proceedings of the 42nd International Conference on Machine Learning},
     year = {2025},
   }

5. Non-Asymptotic Length Generalization

   Thomas Chen, Tengyu Ma, and Zhiyuan Li

   In Proceedings of the 42nd International Conference on Machine Learning, ICML 2025

   arXiv Bib

   @inproceedings{chen2025non,
     author = {Chen, Thomas and Ma, Tengyu and Li, Zhiyuan},
     title = {Non-Asymptotic Length Generalization},
     booktitle = {Proceedings of the 42nd International Conference on Machine Learning},
     year = {2025},
   }

6. Chain-of-Thought Provably Enables Learning the (Otherwise) Unlearnable

   Chenxiao Yang, Zhiyuan Li, and David Wipf

   In The Thirteenth International Conference on Learning Representations, ICLR 2025

   LINK Bib

   @inproceedings{yang2025chain,
     author = {Yang, Chenxiao and Li, Zhiyuan and Wipf, David},
     title = {Chain-of-Thought Provably Enables Learning the (Otherwise) Unlearnable},
     booktitle = {The Thirteenth International Conference on Learning Representations},
     year = {2025},
   }

7. A Coefficient Makes SVRG Effective

   Yida Yin, Zhiqiu Xu, Zhiyuan Li, Trevor Darrell, and Zhuang Liu

   In The Thirteenth International Conference on Learning Representations, ICLR 2025

   arXiv Bib

   @inproceedings{yin2023coefficient,
     author = {Yin, Yida and Xu, Zhiqiu and Li, Zhiyuan and Darrell, Trevor and Liu, Zhuang},
     title = {A Coefficient Makes SVRG Effective},
     booktitle = {The Thirteenth International Conference on Learning Representations},
     year = {2025},
   }

8. Reasoning with Latent Thoughts: On the Power of Looped Transformers

   Nikunj Saunshi, Nishanth Dikkala, Zhiyuan Li, Sanjiv Kumar, and Sashank J Reddi

   In The Thirteenth International Conference on Learning Representations, ICLR 2025

   arXiv Bib

   @inproceedings{saunshi2025reasoning,
     author = {Saunshi, Nikunj and Dikkala, Nishanth and Li, Zhiyuan and Kumar, Sanjiv and Reddi, Sashank J},
     title = {Reasoning with Latent Thoughts: On the Power of Looped Transformers},
     booktitle = {The Thirteenth International Conference on Learning Representations},
     year = {2025},
   }

9. Understanding Warmup-Stable-Decay Learning Rates: A River Valley Loss Landscape Perspective

   Kaiyue Wen, Zhiyuan Li, Jason Wang, David Hall, Percy Liang, and Tengyu Ma

   In The Thirteenth International Conference on Learning Representations, ICLR 2025

   arXiv Bib

   @inproceedings{wen2024understanding,
     author = {Wen, Kaiyue and Li, Zhiyuan and Wang, Jason and Hall, David and Liang, Percy and Ma, Tengyu},
     title = {Understanding Warmup-Stable-Decay Learning Rates: A River Valley Loss Landscape Perspective},
     booktitle = {The Thirteenth International Conference on Learning Representations},
     year = {2025},
   }

10. Adam Exploits $\ell_\infty$-geometry of Loss Landscape via Coordinate-wise Adaptivity

    Shuo Xie, Mohamad Amin Mohamadi, and Zhiyuan Li

    In The Thirteenth International Conference on Learning Representations, ICLR 2025

    Spotlight arXiv Bib

    2025 ICLR Spotlight

    @inproceedings{xie2024adam,
      author = {Xie, Shuo and Mohamadi, Mohamad Amin and Li, Zhiyuan},
      title = {Adam Exploits $\ell_\infty$-geometry of Loss Landscape via Coordinate-wise Adaptivity},
      booktitle = {The Thirteenth International Conference on Learning Representations},
      year = {2025},
    }

2024

1. Implicit Bias of AdamW: $\ell_\infty$-Norm Constrained Optimization

   Shuo Xie and Zhiyuan Li

   In Proceedings of the 41st International Conference on Machine Learning, ICML 2024

   arXiv Bib

   @inproceedings{xie2024implicit,
     author = {Xie, Shuo and Li, Zhiyuan},
     title = {Implicit Bias of AdamW: $\ell_\infty$-Norm Constrained Optimization},
     booktitle = {Proceedings of the 41st International Conference on Machine Learning},
     pages = {54488--54510},
     year = {2024},
     organization = {PMLR},
   }

2. Why Do You Grok? A Theoretical Analysis of Grokking Modular Addition

   Mohamad Amin Mohamadi, Zhiyuan Li, Lei Wu, and Danica J Sutherland

   In Proceedings of the 41st International Conference on Machine Learning, ICML 2024

   arXiv Bib

   @inproceedings{mohamadi2024you,
     author = {Mohamadi, Mohamad Amin and Li, Zhiyuan and Wu, Lei and Sutherland, Danica J},
     title = {Why Do You Grok? A Theoretical Analysis of Grokking Modular Addition},
     booktitle = {Proceedings of the 41st International Conference on Machine Learning},
     pages = {54511--54533},
     year = {2024},
     organization = {PMLR},
   }

3. Simplicity Bias via Global Convergence of Sharpness Minimization

   Khashayar Gatmiry, Zhiyuan Li, Sashank J Reddi, and Stefanie Jegelka

   In International Conference on Machine Learning, ICML 2024

   arXiv Bib

   @inproceedings{gatmiry2024simplicity,
     author = {Gatmiry, Khashayar and Li, Zhiyuan and Reddi, Sashank J and Jegelka, Stefanie},
     title = {Simplicity Bias via Global Convergence of Sharpness Minimization},
     booktitle = {International Conference on Machine Learning},
     pages = {15102--15129},
     year = {2024},
     organization = {PMLR},
   }

4. Chain of Thought Empowers Transformers to Solve Inherently Serial Problems

   Zhiyuan Li, Hong Liu, Denny Zhou, and Tengyu Ma

   In The Twelfth International Conference on Learning Representations, ICLR 2024

   arXiv Bib

   @inproceedings{li2024chain,
     author = {Li, Zhiyuan and Liu, Hong and Zhou, Denny and Ma, Tengyu},
     title = {Chain of Thought Empowers Transformers to Solve Inherently Serial Problems},
     booktitle = {The Twelfth International Conference on Learning Representations},
     year = {2024},
   }

5. Fast Equilibrium of SGD in Generic Situations

   Zhiyuan Li, Yi Wang, and Zhiren Wang (α-β)

   In The Twelfth International Conference on Learning Representations, ICLR 2024

   LINK Bib

   @inproceedings{li2024fast,
     author = {Li, Zhiyuan and Wang, Yi and Wang, Zhiren},
     title = {Fast Equilibrium of SGD in Generic Situations},
     booktitle = {The Twelfth International Conference on Learning Representations},
     year = {2024},
   }

6. Dichotomy of Early and Late Phase Implicit Biases Can Provably Induce Grokking

   Kaifeng Lyu, Jikai Jin, Zhiyuan Li, Simon S Du, Jason D Lee, and Wei Hu

   In The Twelfth International Conference on Learning Representations, ICLR 2024

   arXiv Bib

   @inproceedings{lyu2023dichotomy,
     author = {Lyu, Kaifeng and Jin, Jikai and Li, Zhiyuan and Du, Simon S and Lee, Jason D and Hu, Wei},
     title = {Dichotomy of Early and Late Phase Implicit Biases Can Provably Induce Grokking},
     booktitle = {The Twelfth International Conference on Learning Representations},
     year = {2024},
   }

7. Sophia: A Scalable Stochastic Second-Order Optimizer for Language Model Pre-Training

   Hong Liu, Zhiyuan Li, David Hall, Percy Liang, and Tengyu Ma

   In The Twelfth International Conference on Learning Representations, ICLR 2024

   arXiv Bib

   @inproceedings{liu2023sophia,
     author = {Liu, Hong and Li, Zhiyuan and Hall, David and Liang, Percy and Ma, Tengyu},
     title = {Sophia: A Scalable Stochastic Second-Order Optimizer for Language Model Pre-Training},
     booktitle = {The Twelfth International Conference on Learning Representations},
     year = {2024},
   }

8. The Marginal Value of Momentum for Small Learning Rate SGD

   Runzhe Wang, Sadhika Malladi, Tianhao Wang, Kaifeng Lyu, and Zhiyuan Li

   In The Twelfth International Conference on Learning Representations, ICLR 2024

   arXiv Bib

   @inproceedings{wang2023marginal,
     author = {Wang, Runzhe and Malladi, Sadhika and Wang, Tianhao and Lyu, Kaifeng and Li, Zhiyuan},
     title = {The Marginal Value of Momentum for Small Learning Rate SGD},
     booktitle = {The Twelfth International Conference on Learning Representations},
     year = {2024},
   }

2023

1. What is the Inductive Bias of Flatness Regularization? A Study of Deep Matrix Factorization Models

   Khashayar Gatmiry, Zhiyuan Li, Tengyu Ma, Sashank Reddi, Stefanie Jegelka, and Ching-Yao Chuang

   In Advances in Neural Information Processing Systems 36, NeurIPS 2023

   arXiv Bib

   @inproceedings{gatmiry2023inductive,
     author = {Gatmiry, Khashayar and Li, Zhiyuan and Ma, Tengyu and Reddi, Sashank and Jegelka, Stefanie and Chuang, Ching-Yao},
     title = {What is the Inductive Bias of Flatness Regularization? A Study of Deep Matrix Factorization Models},
     booktitle = {Advances in Neural Information Processing Systems 36},
     volume = {36},
     pages = {28040--28052},
     year = {2023},
   }

2. Sharpness Minimization Algorithms Do Not Only Minimize Sharpness to Achieve Better Generalization

   Kaiyue Wen, Zhiyuan Li, and Tengyu Ma

   In Advances in Neural Information Processing Systems 36, NeurIPS 2023

   Oral arXiv Bib

   2023 NeurIPS Oral

   @inproceedings{wen2023sharpness,
     author = {Wen, Kaiyue and Li, Zhiyuan and Ma, Tengyu},
     title = {Sharpness Minimization Algorithms Do Not Only Minimize Sharpness to Achieve Better Generalization},
     booktitle = {Advances in Neural Information Processing Systems 36},
     volume = {36},
     pages = {1024--1035},
     year = {2023},
   }

3. Understanding Incremental Learning of Gradient Descent: A Fine-Grained Analysis of Matrix Sensing

   Jikai Jin, Zhiyuan Li, Kaifeng Lyu, Simon Shaolei Du, and Jason D Lee

   In Proceedings of the 40th International Conference on Machine Learning, ICML 2023

   arXiv Bib

   @inproceedings{jin2023understanding,
     author = {Jin, Jikai and Li, Zhiyuan and Lyu, Kaifeng and Du, Simon Shaolei and Lee, Jason D},
     title = {Understanding Incremental Learning of Gradient Descent: A Fine-Grained Analysis of Matrix Sensing},
     booktitle = {Proceedings of the 40th International Conference on Machine Learning},
     pages = {15200--15238},
     year = {2023},
     organization = {PMLR},
   }

4. Same Pre-Training Loss, Better Downstream: Implicit Bias Matters for Language Models

   Hong Liu, Sang Michael Xie, Zhiyuan Li, and Tengyu Ma

   In Proceedings of the 40th International Conference on Machine Learning, ICML 2023

   arXiv Bib

   @inproceedings{liu2023same,
     author = {Liu, Hong and Xie, Sang Michael and Li, Zhiyuan and Ma, Tengyu},
     title = {Same Pre-Training Loss, Better Downstream: Implicit Bias Matters for Language Models},
     booktitle = {Proceedings of the 40th International Conference on Machine Learning},
     pages = {22188--22214},
     year = {2023},
     organization = {PMLR},
   }

5. How Does Sharpness-Aware Minimization Minimize Sharpness?

   Kaiyue Wen, Tengyu Ma, and Zhiyuan Li

   In The Eleventh International Conference on Learning Representations, ICLR 2023

   arXiv Bib

   @inproceedings{wen2023does,
     author = {Wen, Kaiyue and Ma, Tengyu and Li, Zhiyuan},
     title = {How Does Sharpness-Aware Minimization Minimize Sharpness?},
     booktitle = {The Eleventh International Conference on Learning Representations},
     year = {2023},
   }

2022

1. Fast Mixing of Stochastic Gradient Descent with Normalization and Weight Decay

   Zhiyuan Li, Tianhao Wang, and Dingli Yu

   In Advances in Neural Information Processing Systems 35, NeurIPS 2022

   LINK Bib

   @inproceedings{li2022fast,
     author = {Li, Zhiyuan and Wang, Tianhao and Yu, Dingli},
     title = {Fast Mixing of Stochastic Gradient Descent with Normalization and Weight Decay},
     booktitle = {Advances in Neural Information Processing Systems 35},
     volume = {35},
     pages = {9233--9248},
     year = {2022},
   }

2. Implicit Bias of Gradient Descent on Reparametrized Models: On Equivalence to Mirror Descent

   Zhiyuan Li^*, Tianhao Wang^*, Jason D Lee, and Sanjeev Arora

   In Advances in Neural Information Processing Systems 35, NeurIPS 2022

   arXiv Bib

   @inproceedings{li2022implicit,
     author = {Li, Zhiyuan and Wang, Tianhao and Lee, Jason D and Arora, Sanjeev},
     title = {Implicit Bias of Gradient Descent on Reparametrized Models: On Equivalence to Mirror Descent},
     booktitle = {Advances in Neural Information Processing Systems 35},
     volume = {35},
     pages = {34626--34640},
     year = {2022},
   }

3. Understanding the Generalization Benefit of Normalization Layers: Sharpness Reduction

   Kaifeng Lyu, Zhiyuan Li, and Sanjeev Arora

   In Advances in Neural Information Processing Systems 35, NeurIPS 2022

   arXiv Bib

   @inproceedings{lyu2022understanding,
     author = {Lyu, Kaifeng and Li, Zhiyuan and Arora, Sanjeev},
     title = {Understanding the Generalization Benefit of Normalization Layers: Sharpness Reduction},
     booktitle = {Advances in Neural Information Processing Systems 35},
     volume = {35},
     pages = {34689--34708},
     year = {2022},
   }

4. Bridging Theory and Practice in Deep Learning: Optimization and Generalization

   Zhiyuan Li

   Princeton University, PhD Thesis 2022

   Bib

   @phdthesis{li2022bridging,
     author = {Li, Zhiyuan},
     title = {Bridging Theory and Practice in Deep Learning: Optimization and Generalization},
     year = {2022},
     school = {Princeton University},
   }

5. Understanding Gradient Descent on the Edge of Stability in Deep Learning

   Sanjeev Arora, Zhiyuan Li, and Abhishek Panigrahi (α-β)

   In Proceedings of the 39th International Conference on Machine Learning, ICML 2022

   arXiv Bib

   @inproceedings{arora2022understanding,
     author = {Arora, Sanjeev and Li, Zhiyuan and Panigrahi, Abhishek},
     title = {Understanding Gradient Descent on the Edge of Stability in Deep Learning},
     booktitle = {Proceedings of the 39th International Conference on Machine Learning},
     pages = {948--1024},
     year = {2022},
     organization = {PMLR},
   }

6. Robust Training of Neural Networks Using Scale Invariant Architectures

   Zhiyuan Li, Srinadh Bhojanapalli, Manzil Zaheer, Sashank Reddi, and Sanjiv Kumar

   In Proceedings of the 39th International Conference on Machine Learning, ICML 2022

   Long Presentation arXiv Bib

   2022 ICML Long Presentation

   @inproceedings{li2022robust,
     author = {Li, Zhiyuan and Bhojanapalli, Srinadh and Zaheer, Manzil and Reddi, Sashank and Kumar, Sanjiv},
     title = {Robust Training of Neural Networks Using Scale Invariant Architectures},
     booktitle = {Proceedings of the 39th International Conference on Machine Learning},
     pages = {12656--12684},
     year = {2022},
     organization = {PMLR},
   }

7. What Happens after SGD Reaches Zero Loss?--A Mathematical Framework

   Zhiyuan Li, Tianhao Wang, and Sanjeev Arora

   In The Tenth International Conference on Learning Representations, ICLR 2022

   Spotlight arXiv Bib

   2022 ICLR Spotlight

   @inproceedings{li2021happens,
     author = {Li, Zhiyuan and Wang, Tianhao and Arora, Sanjeev},
     title = {What Happens after SGD Reaches Zero Loss?--A Mathematical Framework},
     booktitle = {The Tenth International Conference on Learning Representations},
     year = {2022},
   }

2021

1. Gradient Descent on Two-Layer Nets: Margin Maximization and Simplicity Bias

   Kaifeng Lyu^*, Zhiyuan Li^*, Runzhe Wang^*, and Sanjeev Arora

   In Advances in Neural Information Processing Systems 34, NeurIPS 2021

   arXiv Bib

   @inproceedings{lyu2021gradient,
     author = {Lyu, Kaifeng and Li, Zhiyuan and Wang, Runzhe and Arora, Sanjeev},
     title = {Gradient Descent on Two-Layer Nets: Margin Maximization and Simplicity Bias},
     booktitle = {Advances in Neural Information Processing Systems 34},
     volume = {34},
     pages = {12978--12991},
     year = {2021},
   }

2. When Is Particle Filtering Efficient For Planning In Partially Observed Linear Dynamical Systems?

   Simon S Du, Wei Hu, Zhiyuan Li, Ruoqi Shen, Zhao Song, and Jiajun Wu

   In Proceedings of the 37th Conference on Uncertainty in Artificial Intelligence, UAI 2021

   arXiv Bib

   @inproceedings{du2021particle,
     author = {Du, Simon S and Hu, Wei and Li, Zhiyuan and Shen, Ruoqi and Song, Zhao and Wu, Jiajun},
     title = {When Is Particle Filtering Efficient For Planning In Partially Observed Linear Dynamical Systems?},
     booktitle = {Proceedings of the 37th Conference on Uncertainty in Artificial Intelligence},
     pages = {728--737},
     year = {2021},
     organization = {PMLR},
   }

3. Risk Bounds And Rademacher Complexity In Batch Reinforcement Learning

   Yaqi Duan, Chi Jin, and Zhiyuan Li (α-β)

   In Proceedings of the 38th International Conference on Machine Learning, ICML 2021

   arXiv Bib

   @inproceedings{duan2021risk,
     author = {Duan, Yaqi and Jin, Chi and Li, Zhiyuan},
     title = {Risk Bounds And Rademacher Complexity In Batch Reinforcement Learning},
     booktitle = {Proceedings of the 38th International Conference on Machine Learning},
     pages = {2892--2902},
     year = {2021},
     organization = {PMLR},
   }

4. On The Validity Of Modeling Sgd With Stochastic Differential Equations (SDEs)

   Zhiyuan Li, Sadhika Malladi, and Sanjeev Arora

   In Advances in Neural Information Processing Systems 34, NeurIPS 2021

   arXiv Bib

   @inproceedings{li2021validity,
     author = {Li, Zhiyuan and Malladi, Sadhika and Arora, Sanjeev},
     title = {On The Validity Of Modeling Sgd With Stochastic Differential Equations (SDEs)},
     booktitle = {Advances in Neural Information Processing Systems 34},
     volume = {34},
     pages = {12712--12725},
     year = {2021},
   }

5. Towards Resolving The Implicit Bias Of Gradient Descent For Matrix Factorization: Greedy Low-Rank Learning

   Zhiyuan Li, Yuping Luo, and Kaifeng Lyu (α-β)

   In The Ninth International Conference on Learning Representations, ICLR 2021

   arXiv Bib

   @inproceedings{li2020towards,
     author = {Li, Zhiyuan and Luo, Yuping and Lyu, Kaifeng},
     title = {Towards Resolving The Implicit Bias Of Gradient Descent For Matrix Factorization: Greedy Low-Rank Learning},
     booktitle = {The Ninth International Conference on Learning Representations},
     year = {2021},
   }

6. Why Are Convolutional Nets More Sample-Efficient Than Fully-Connected Nets?

   Zhiyuan Li, Yi Zhang, and Sanjeev Arora

   In The Ninth International Conference on Learning Representations, ICLR 2021

   Oral arXiv Bib

   2021 ICLR Oral

   @inproceedings{li2020convolutional,
     author = {Li, Zhiyuan and Zhang, Yi and Arora, Sanjeev},
     title = {Why Are Convolutional Nets More Sample-Efficient Than Fully-Connected Nets?},
     booktitle = {The Ninth International Conference on Learning Representations},
     year = {2021},
   }

2020

1. Implicit Regularization And Convergence For Weight Normalization

   Xiaoxia Wu, Edgar Dobriban, Tongzheng Ren, Shanshan Wu, Zhiyuan Li, Suriya Gunasekar, Rachel Ward, and Qiang Liu

   In Advances in Neural Information Processing Systems 33, NeurIPS 2020

   arXiv Bib

   @inproceedings{wu2020implicit,
     author = {Wu, Xiaoxia and Dobriban, Edgar and Ren, Tongzheng and Wu, Shanshan and Li, Zhiyuan and Gunasekar, Suriya and Ward, Rachel and Liu, Qiang},
     title = {Implicit Regularization And Convergence For Weight Normalization},
     booktitle = {Advances in Neural Information Processing Systems 33},
     volume = {33},
     pages = {2835--2847},
     year = {2020},
   }

2. Reconciling Modern Deep Learning With Traditional Optimization Analyses: The Intrinsic Learning Rate

   Zhiyuan Li^*, Kaifeng Lyu^*, and Sanjeev Arora

   In Advances in Neural Information Processing Systems 33, NeurIPS 2020

   arXiv Bib

   @inproceedings{li2020reconciling,
     author = {Li, Zhiyuan and Lyu, Kaifeng and Arora, Sanjeev},
     title = {Reconciling Modern Deep Learning With Traditional Optimization Analyses: The Intrinsic Learning Rate},
     booktitle = {Advances in Neural Information Processing Systems 33},
     volume = {33},
     pages = {14544--14555},
     year = {2020},
   }

3. Simple and Effective Regularization Methods for Training on Noisily Labeled Data with Generalization Guarantee

   Wei Hu, Zhiyuan Li, and Dingli Yu

   In The Eighth International Conference on Learning Representations, ICLR 2020

   arXiv Bib

   @inproceedings{hu2019simple,
     author = {Hu, Wei and Li, Zhiyuan and Yu, Dingli},
     title = {Simple and Effective Regularization Methods for Training on Noisily Labeled Data with Generalization Guarantee},
     booktitle = {The Eighth International Conference on Learning Representations},
     year = {2020},
   }

4. Harnessing the power of infinitely wide deep nets on small-data tasks

   Sanjeev Arora, Simon S Du, Zhiyuan Li, Ruslan Salakhutdinov, Ruosong Wang, and Dingli Yu (α-β)

   In The Eighth International Conference on Learning Representations, ICLR 2020

   Spotlight arXiv Bib

   2020 ICLR Spotlight

   @inproceedings{arora2019harnessing,
     author = {Arora, Sanjeev and Du, Simon S and Li, Zhiyuan and Salakhutdinov, Ruslan and Wang, Ruosong and Yu, Dingli},
     title = {Harnessing the power of infinitely wide deep nets on small-data tasks},
     booktitle = {The Eighth International Conference on Learning Representations},
     year = {2020},
   }

5. An Exponential Learning Rate Schedule for Deep Learning

   Zhiyuan Li and Sanjeev Arora

   In The Eighth International Conference on Learning Representations, ICLR 2020

   Spotlight arXiv Bib

   2020 ICLR Spotlight

   @inproceedings{li2019exponential,
     author = {Li, Zhiyuan and Arora, Sanjeev},
     title = {An Exponential Learning Rate Schedule for Deep Learning},
     booktitle = {The Eighth International Conference on Learning Representations},
     year = {2020},
   }

2019

1. Enhanced Convolutional Neural Tangent Kernels

   Zhiyuan Li^*, Ruosong Wang^*, Dingli Yu^*, Simon S Du, Wei Hu, Ruslan Salakhutdinov, and Sanjeev Arora

   In arXiv preprint arXiv:1911.00809, arXiv 2019

   arXiv Bib

   @inproceedings{li2019enhanced,
     author = {Li, Zhiyuan and Wang, Ruosong and Yu, Dingli and Du, Simon S and Hu, Wei and Salakhutdinov, Ruslan and Arora, Sanjeev},
     title = {Enhanced Convolutional Neural Tangent Kernels},
     booktitle = {arXiv preprint arXiv:1911.00809},
     year = {2019},
   }

2. Explaining Landscape Connectivity of Low-Cost Solutions for Multilayer Nets

   Rohith Kuditipudi, Xiang Wang, Holden Lee, Yi Zhang, Zhiyuan Li, Wei Hu, Rong Ge, and Sanjeev Arora

   In Advances in Neural Information Processing Systems 32, NeurIPS 2019

   arXiv Bib

   @inproceedings{kuditipudi2019explaining,
     author = {Kuditipudi, Rohith and Wang, Xiang and Lee, Holden and Zhang, Yi and Li, Zhiyuan and Hu, Wei and Ge, Rong and Arora, Sanjeev},
     title = {Explaining Landscape Connectivity of Low-Cost Solutions for Multilayer Nets},
     booktitle = {Advances in Neural Information Processing Systems 32},
     volume = {32},
     year = {2019},
   }

3. On Exact Computation with an Infinitely Wide Neural Net

   Sanjeev Arora, Simon S Du, Wei Hu, Zhiyuan Li, Russ R Salakhutdinov, and Ruosong Wang (α-β)

   In Advances in Neural Information Processing Systems 32, NeurIPS 2019

   Spotlight arXiv Bib

   2019 NeurIPS Spotlight

   @inproceedings{arora2019exact,
     author = {Arora, Sanjeev and Du, Simon S and Hu, Wei and Li, Zhiyuan and Salakhutdinov, Russ R and Wang, Ruosong},
     title = {On Exact Computation with an Infinitely Wide Neural Net},
     booktitle = {Advances in Neural Information Processing Systems 32},
     volume = {32},
     year = {2019},
   }

4. Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks

   Sanjeev Arora, Simon Du, Wei Hu, Zhiyuan Li, and Ruosong Wang (α-β)

   In Proceedings of the 36th International Conference on Machine Learning, ICML 2019

   arXiv Bib

   @inproceedings{arora2019fine,
     author = {Arora, Sanjeev and Du, Simon and Hu, Wei and Li, Zhiyuan and Wang, Ruosong},
     title = {Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks},
     booktitle = {Proceedings of the 36th International Conference on Machine Learning},
     pages = {322--332},
     year = {2019},
     organization = {PMLR},
   }

5. Theoretical Analysis of Auto Rate-Tuning by Batch Normalization

   Sanjeev Arora, Zhiyuan Li, and Kaifeng Lyu (α-β)

   In The Seventh International Conference on Learning Representations, ICLR 2019

   arXiv Bib

   @inproceedings{arora2018theoretical,
     author = {Arora, Sanjeev and Li, Zhiyuan and Lyu, Kaifeng},
     title = {Theoretical Analysis of Auto Rate-Tuning by Batch Normalization},
     booktitle = {The Seventh International Conference on Learning Representations},
     year = {2019},
   }

6. Towards Understanding the Role of Over-Parametrization in Generalization of Neural Networks

   Behnam Neyshabur, Zhiyuan Li, Srinadh Bhojanapalli, Yann LeCun, and Nathan Srebro

   In The Seventh International Conference on Learning Representations, ICLR 2019

   arXiv Bib

   @inproceedings{neyshabur2018towards,
     author = {Neyshabur, Behnam and Li, Zhiyuan and Bhojanapalli, Srinadh and LeCun, Yann and Srebro, Nathan},
     title = {Towards Understanding the Role of Over-Parametrization in Generalization of Neural Networks},
     booktitle = {The Seventh International Conference on Learning Representations},
     year = {2019},
   }

2018

1. Online Improper Learning with an Approximation Oracle

   Elad Hazan, Wei Hu, Yuanzhi Li, and Zhiyuan Li (α-β)

   In Advances in Neural Information Processing Systems 31, NeurIPS 2018

   arXiv Bib

   @inproceedings{hazan2018online,
     author = {Hazan, Elad and Hu, Wei and Li, Yuanzhi and Li, Zhiyuan},
     title = {Online Improper Learning with an Approximation Oracle},
     booktitle = {Advances in Neural Information Processing Systems 31},
     pages = {5657--5665},
     year = {2018},
   }

2017

1. Stability of Generalized Two-Sided Markets with Transaction Thresholds

   Zhiyuan Li, Yicheng Liu, Pingzhong Tang, Tingting Xu, and Wei Zhan (α-β)

   In Proceedings of the 16th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2017

   Best Paper Nomination LINK Bib

   2017 AAMAS Best Paper Nomination

   @inproceedings{li2017stability,
     author = {Li, Zhiyuan and Liu, Yicheng and Tang, Pingzhong and Xu, Tingting and Zhan, Wei},
     title = {Stability of Generalized Two-Sided Markets with Transaction Thresholds},
     booktitle = {Proceedings of the 16th International Conference on Autonomous Agents and Multiagent Systems},
     pages = {290--298},
     year = {2017},
   }

2016

1. Learning in Games: Robustness of Fast Convergence

   Dylan J Foster, Zhiyuan Li, Thodoris Lykouris, Karthik Sridharan, and Eva Tardos (α-β)

   In Advances in Neural Information Processing Systems 29, NeurIPS 2016

   arXiv Bib

   @inproceedings{foster2016learning,
     author = {Foster, Dylan J and Li, Zhiyuan and Lykouris, Thodoris and Sridharan, Karthik and Tardos, Eva},
     title = {Learning in Games: Robustness of Fast Convergence},
     booktitle = {Advances in Neural Information Processing Systems 29},
     volume = {29},
     year = {2016},
   }

2. Solving Marginal MAP Problems with NP Oracles and Parity Constraints

   Yexiang Xue, Zhiyuan Li, Stefano Ermon, Carla P Gomes, and Bart Selman

   In Advances in Neural Information Processing Systems 29, NeurIPS 2016

   arXiv Bib

   @inproceedings{xue2016solving,
     author = {Xue, Yexiang and Li, Zhiyuan and Ermon, Stefano and Gomes, Carla P and Selman, Bart},
     title = {Solving Marginal MAP Problems with NP Oracles and Parity Constraints},
     booktitle = {Advances in Neural Information Processing Systems 29},
     volume = {29},
     year = {2016},
   }