Publications
See also my Google Scholar profile. * denotes equal contribution; in the TKDE paper, the two co-first authors are listed with Gang Wen first. ** denotes alphabetical order.
Conference and workshop papers
-
Fast Learning Rate Transfer for Gradient Descent in Sketched Linear Regression. G. Wen, A. Bietti, N. Ghosh, T. Misiakiewicz, D. Wu. ICML Workshop on High-Dimensional Learning Dynamics (HiLD), spotlight (2026). Preliminary version of On fast hyperparameter transfer under fixed and diverging time horizons.
-
PeriodicSketch: Finding Periodic Items in Data Streams. Z. Fan, Y. Zhang, T. Yang, M. Yan, G. Wen, Y. Wu, H. Li, B. Cui. 2022 IEEE 38th International Conference on Data Engineering (ICDE), 96–109 (2022).
Journal papers
-
Recommending Best Paper Awards via the Isotonic Mechanism. G. G. Wen, B. Su, N. Collina, Z. Deng, W. Su. Accepted for publication in Journal of Machine Learning Research (2026).
-
Self-rankings as a predictor of scientific impact beyond peer review. B. Su, N. Collina, G. Wen, D. Li, K. Cho, J. Fan, B. Zhao, W. Su. Nature Computational Science (2026).
-
On the Evolutionary of Bloom Filter False Positives — An Information Theoretical Approach to Optimizing Bloom Filter Parameters. G. Wen*, Z. Fan*, Z. Huang, Y. Zhou, Q. Fu, T. Yang, A. X. Liu, B. Cui. IEEE Transactions on Knowledge and Data Engineering, 35(7), 7316–7327 (2023).
Preprints and technical reports
-
textGrain: Entropy-Calibrated Watermarking for Language Model Text. X. Li, G. Wen, X. Chen, Q. Long, A. Jain, F. Joly, M. Lam, Q. Song, W. Su. Technical report accompanying OpenAI's October 5, 2026 announcement (2026).
-
On fast hyperparameter transfer under fixed and diverging time horizons. G. G. Wen, A. Bietti, N. Ghosh, T. Misiakiewicz, D. Wu. Working paper; under review (2026).
-
Optimal Estimation of Watermark Proportions in Hybrid AI-Human Texts. X. Li, G. Wen, W. He, J. Wu, Q. Long, W. J. Su. Minor revision at Journal of the American Statistical Association; arXiv:2506.22343 (2025).
-
The sharp SAT/UNSAT phase transition in random ellipsoid fitting. T. Misiakiewicz**, G. G. Wen**. Submitted to The Annals of Probability; arXiv:2608.10184 (2026).
-
Dynamical mean-field limit and replica-symmetric free energy for the orthogonally-invariant SK model. Z. Fan**, T. Misiakiewicz**, L. Wang**, G. G. Wen**. Submitted to The Annals of Probability; arXiv:2607.10102 (2026).
-
When does Gaussian equivalence fail and how to fix it: Non-universal behavior of random features with quadratic scaling. G. G. Wen, H. Hu, Y. M. Lu, Z. Fan, T. Misiakiewicz. Submitted to Mathematical Statistics and Learning; arXiv:2512.03325 (2025).
-
Optimal Detection for Language Watermarks with Pseudorandom Collision. T. Tony Cai**, X. Li**, Q. Long**, W. J. Su**, G. G. Wen**. arXiv:2510.22007 (2025).