Publications

The asterisks mean that authors are sorted in alphabetical order.

2025

  1. ArXiv Preprint
    Not All Rollouts are Useful: Down-Sampling Rollouts in LLM Reinforcement Learning
    Yixuan Even Xu*, Yash Savani*, Fei Fang, and J. Zico Kolter
    arXiv preprint arXiv:2504.13818, 2025
    Talk invited at Bosch Research Pittsburgh
    arXiv PDF Slides
  2. ArXiv Preprint
    Existing Large Language Model Unlearning Evaluations Are Inconclusive
    Zhili Feng*, Yixuan Even Xu*, Alexander Robey, Robert Kirk, Xander Davies, Yarin Gal, Avi Schwarzschild, and J. Zico Kolter
    arXiv preprint arXiv:2506.00688, 2025
    arXiv PDF
  3. NeurIPS 2025
    Antidistillation Sampling
    Yash Savani*, Asher Trockman*, Zhili Feng, Yixuan Even Xu, Avi Schwarzschild, Alexander Robey, Marc Finzi, and J. Zico Kolter
    In Advances in Neural Information Processing Systems, 2025
    Talk invited at OpenAI, Google and Simons Institute
    arXiv PDF
  4. Workshop
    Evaluating LLM Memorization Using Soft Token Sparsity
    Zhili Feng*, Yixuan Even Xu*, Pratyush Maini, Alexander Robey, Avi Schwarzschild, and J. Zico Kolter
    In ICLR Workshop on Sparsity in LLMs, 2025
    PDF
  5. AAAI 2025
    Deviate or Not: Learning Coalition Structures with Multiple-bit Observations in Games
    Yixuan Even Xu, Zhe Feng, and Fei Fang
    In Proceedings of the AAAI Conference on Artificial Intelligence, 2025
    arXiv PDF Poster Slides Video
  6. AAAI 2025
    Improving Community-Participated Patrol for Anti-Poaching
    Yufei Wu, Yixuan Even Xu, Zhang Xuming, Liu Duo, Zhu Shibing, and Fei Fang
    In Proceedings of the AAAI Conference on Artificial Intelligence, 2025
    Selected for Oral Presentation
    arXiv PDF Poster Slides Video

2024

  1. NeurIPS 2024
    Aggregating Quantitative Relative Judgments: From Social Choice to Ranking Prediction
    Yixuan Even Xu, Hanrui Zhang, Yu Cheng, and Vincent Conitzer
    In Advances in Neural Information Processing Systems, 2024
    arXiv PDF Poster Slides Video
  2. WWW 2024
    Bidder Selection Problem in Position Auctions: A Fast and Simple Algorithm via Poisson Approximation
    Nick Gravin*, Yixuan Even Xu*, and Renfei Zhou*
    In Proceedings of the ACM Web Conference, 2024
    Selected for Oral Presentation
    arXiv PDF Poster Slides Video
  3. AAAI 2024
    Learning Coalition Structures with Games
    Yixuan Even Xu, Chun Kai Ling, and Fei Fang
    In Proceedings of the AAAI Conference on Artificial Intelligence, 2024
    arXiv PDF Poster Slides Video
  4. AAAI 2024
    Non-excludable Bilateral Trade between Groups
    Yixuan Even Xu, Hanrui Zhang, and Vincent Conitzer
    In Proceedings of the AAAI Conference on Artificial Intelligence, 2024
    arXiv PDF Poster Slides Video

2023

  1. NeurIPS 2023
    A One-Size-Fits-All Approach to Improving Randomness in Paper Assignment
    Yixuan Even Xu, Steven Jecmen, Zimeng Song, and Fei Fang
    In Advances in Neural Information Processing Systems, 2023
    Selected for Spotlight Presentation
    Deployed in NeurIPS 2024 and 2025
    Covered by NeurIPS conference blog
    Subsequent work deployed in AAAI 2026 and AAMAS 2026
    arXiv PDF Poster Slides Video
  2. ESA 2023
    On the Perturbation Function of Ranking and Balance for Weighted Online Bipartite Matching
    Jingxun Liang*, Zhihao Gavin Tang*, Yixuan Even Xu*, Yuhao Zhang*, and Renfei Zhou*
    In European Symposium on Algorithms, 2023
    arXiv PDF