报告题目：Efficient Online LLM Watermark Detection via Rao–Blackwellized E-Processes

主讲人：李婷副教授（上海财经大学）

时间：2026年10月27日（周二）14:00 p.m.

形式：线上讲座

腾讯会议：794-728-270

主办单位：统计与数学学院

摘要：

As large language models (LLMs) are increasingly deployed, reliable and efficient mechanisms for distinguishing AI-generated text from human-written content have become essential. Statistical watermarking has emerged as a promising solution, yet most existing methods are typically fixed-horizon procedures, precluding valid early stopping in streaming generation. In this paper, we develop an efficient online watermark detection framework with anytime-valid inference based on Rao--Blackwellized e-processes, enabling recursive token-level evidence updates without storing the full history. In particular, we instantiate the framework for the Gumbel-max watermark and reduce the original token-level dependence testing problem to a pivot-induced sequential testing problem with an explicit null distribution. Theoretically, we prove anytime-valid Type I error control under arbitrary optional stopping and establish positive asymptotic log-growth under watermarking, implying consistency of the proposed stopping rules. Simulations and experiments on real LLM-generated text demonstrate efficient online detection with rigorous anytime-valid guarantees.

主讲人简介：

李婷，上海财经大学统计与数据科学学院副教授，研究方向为函数型数据分析、医学基因影像数据分析和因果推断。在统计与机器学习权威期刊JASA, AOAS， JMLR上发表过多篇论文。