Yang Li (郦洋)

Master Student @ ReThinkLab, SJTU.

prof_pic.jpg

Mail: yanglily@sjtu.edu.cn

Tel: (+86)15201962556

City: Shanghai, 200240

About

I am now a second-year master student at Department of Computer Science and Engineering from Shanghai Jiao Tong University, supervised by Prof. Junchi Yan. I achieved the Bachelor degree from SJTU in 2022. My research interests lie in machine learning, especially deep generative models and combinatorial optimization on graphs.

I serve as the reviewer for top-tier ML conferences (NeurIPS, ICML, ICLR) and journals (Machine Learning).

Academic Performance

Undergraduate period:

  • GPA: 91.03/100 (or 3.93/4.3), Rank: 3/129 (top 2.3%)
  • Courses: 76.7% above A, 45.0% above A+

Postgraduate period:

  • GPA: 3.83/4.0
  • Courses: 90.0% at A level

Selected Awards

  • Undergraduate National Scholarship (top 0.2% in the nation)
  • Graduate National Scholarship (top 1% in CS Dept.)
  • Outstanding Graduate of Shanghai (top 3%)
  • Huawei Fellowship (top 5%)
  • HyperGryph Fellowship (top 3%)
  • 1st-Class Academic Excellence Scholarship (top 1%)
  • Merit Student of Shanghai Jiao Tong University
  • 1st-Class Academic Scholarship for Graduate Students
  • Special Prize for Social Practice of SJTU
  • First Prize for Social Practice of SJTU
  • Advanced Individuals in Social Practice of SJTU

Selected Projects

  • International SAT Competition 2022 & Solving Strategy Integration to Huawei’s Practical Applications
    The solver Kissat_Adaptive_Restart that I was personally responsible for achieved 12th place in Anniversary Track and 26th place in Main Track worldwide, collaborated with Decision Making and Reasoning Lab, Huawei Noah’s Ark Lab. The solving strategy was integrated into Huawei Hisilicon’s practical applications, with an average performance gain of around 15% and a maximum performance improvement of 97%.

  • Data Strategy Integration to HUAWEI CLOUD’s OptVerse AI Solver
    The computational hardness maintaining data augmentation strategy proposed in our KDD’23 HardSATGEN has been integrated into HUAWEI CLOUD’s OptVerse AI Solver. OptVerse AI Solver solves problems with hundreds of millions of variables, at 100x computing speed thanks to distributed parallel acceleration. OptVerse AI Solver ranks first on the Hans Mittelmann Benchmark for Simplex LP solvers and won the most prestigious award at the World Artificial Intelligence Conference 2023: SAIL (Super AI Leader) Award.

  • Awesome Machine Learning for Combinatorial Optimization Resources
    We maintain a list of resources that utilize machine learning technologies to solve combinatorial optimization problems in awesome-ml4co. The repository covers the learning-based efforts for 34 different combinatorial optimization problems and includes over 300 papers in the ML4CO community. The repository has gained over 1,000 stars on Github.

  • Anime-Paintbrush: A Web-based Automatic Animation Coloring Tool
    A web tool that would automatically assist in generating high quality images by simply inputing a sketch with a little color cue. This project deploys the trained generative adversarial networks (GAN) on the backend, while the frontend is deployed on a web application, and a user-friendly interface is designed so that users can complete the color rendering of sketches on the web page. See Project Page.

News

Jan 16, 2024 One paper was accepted by ICLR 2024.
Sep 25, 2023 I was awarded the Graduate National Scholarship (3/211 in CS Dept.).
Sep 22, 2023 One paper was accepted by NeurIPS 2023.
Jun 11, 2023 I was invited to give a talk on VALSE 2023.
May 17, 2023 One paper was accepted by SIGKDD 2023.

Publications

  1. ICLR
    MixSATGEN: Learning Graph Mixing for SAT Instance Generation
    Xinyan Chen*, Yang Li*, Runzhong Wang, and 1 more author
    In The Twelfth International Conference on Learning Representations, 2024
  2. arXiv
    Machine Learning Insides OptVerse AI Solver: Design Principles and Applications
    Xijun Li, Fangzhou Zhu, Hui-Ling Zhen, and 23 more authors
    arXiv preprint, 2024
  3. arXiv
    Molecule Generation for Drug Design: a Graph Learning Perspective
    Nianzu Yang, Huaijin Wu, Kaipeng Zeng, and 2 more authors
    arXiv preprint, 2024
  4. arXiv
    Rethinking and Benchmarking Predict-then-Optimize Paradigm for Combinatorial Optimization Problems
    Haoyu Geng, Han Ruan, Runzhong Wang, and 4 more authors
    arXiv preprint, 2023
  5. NeurIPS
    T2T: From Distribution Learning in Training to Gradient Search in Testing for Combinatorial Optimization
    Yang Li, Jinpei Guo, Runzhong Wang, and 1 more author
    In Advances in Neural Information Processing Systems, 2023
  6. SIGKDD
    HardSATGEN: Understanding the Difficulty of Hard SAT Formula Generation and A Strong Structure-Hardness-Aware Baseline
    Yang Li, Xinyan Chen, Wenxuan Guo, and 6 more authors
    In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2023
  7. IJCAI
    IID-GAN: an IID Sampling Perspective for Regularizing Mode Collapse
    Yang Li*, Liangliang Shi*, and Junchi Yan
    In Proceedings of the 32nd International Joint Conference on Artificial Intelligence, 2023
  8. NeurIPSSpotlight (top 5%)
    Improving Generative Adversarial Networks via Adversarial Learning in Latent Space
    Yang Li, Yichuan Mo, Liangliang Shi, and 1 more author
    In Advances in Neural Information Processing Systems, 2022
  9. NeurIPS
    The Policy-gradient Placement and Generative Routing Neural Networks for Chip Design
    Ruoyu Cheng, Xianglong Lyu, Yang Li, and 3 more authors
    In Advances in Neural Information Processing Systems, 2022
  10. Kissat Adaptive Restart, Kissat Cfexp: Adaptive Restart Policy and Variable Scoring Improvement
    Yang Li, Yuqi Jia, Wanqian Luo, and 4 more authors
    Proceedings of SAT Competition, 2022