吕凯风 Kaifeng Lyu
我现在是加州大学伯克利分校的Simons研究所的一名博士后研究员，参与项目 Modern Paradigms in Generalization 及 Special Year on Large Language Models and Transformers。我于 2024 年获得普林斯顿大学计算机科学博士学位，师从 Prof. Sanjeev Arora。 本科就读于清华大学姚班，于2019年毕业并取得计算机科学与技术工学学士学位。本科时的学术研究曾由 李建教授 指导。
预印本论文
会议论文
Understanding incremental learning of gradient descent: A fine-grained analysis of matrix sensing
Understanding the Generalization Benefit of Normalization Layers: Sharpness Reduction
New Definitions and Evaluations for Saliency Methods: Staying Intrinsic, Complete and Sound
Towards Resolving the Implicit Bias of Gradient Descent for Matrix Factorization: Greedy Low-Rank Learning
（按字母序排序）
Reconciling Modern Deep Learning with Traditional Optimization Analyses: The Intrinsic Learning Rate
Single-Source Bottleneck Path Algorithm Faster than Sorting for Sparse Graphs
（按字母序排序）
（默认按贡献排序；星号 * 表示贡献相同）
- 普林斯顿大学 2024 年春季学期. Teaching Assistant for COS324: Introduction to Machine Learning (taught by Prof. Sanjeev Arora & Prof. Elad Hazan).
- 普林斯顿大学 2022 年秋季学期. Teaching Assistant for COS521: Advanced Algorithm Design (taught by Prof. Matt Weinberg & Prof. Huacheng Yu).
- 普林斯顿大学 2021 年春季学期. Teaching Assistant for COS598B: Advanced Topics in Computer Science: Mathematical Understanding of Deep Learning (taught by Prof. Sanjeev Arora).
- 清华大学 2020 年春季学期. 计算机应用数学 助教（授课教师：姚期智教授）.
- 清华大学 2019 年春季学期. 分布式计算：基础与系统 助教（授课教师：陈卫教授）.
Professional Services
- Organizer, NeurIPS 2024 Workshop on Mathematics of Modern Machine Learning (M3L 2024).
- Organizer, NeurIPS 2023 Workshop on Mathematics of Modern Machine Learning (M3L).
- Conference Reviewer: ICML (2020-2023), NeurIPS (2020-2023), ICLR (2022-2024), TPAMI, COLT (2020), AAAI (2020), KDD (2022).
- Journal Reviewer: TMLR, JMLR, TPAMI, AIJ.
- Organizer, Yao Class Seminar, Tsinghua University (Fall 2019, Fall 2020, Spring 2021).