Wenya Xie

I’m Wenya Xie, a Ph.D. student in Computer Science at the University of Minnesota (UMN), advised by Prof. Zirui Liu. Previously, I worked as a research assistant at CUHK (Shenzhen) with Prof. Benyou Wang and Feng Jiang. I received my bachelor’s degree from Wuhan University.

My research centers on large language models, spanning long-horizon tasks, reasoning, memory, and human–AI collaboration. Currently I focus on two directions:

• Reasoning & overthinking — why reasoning models waste decoding budget and how they can reason more efficiently and proactively: Word Salad Chopper (EMNLP 2025 Oral) and Reasoning While Asking (ACL 2026).

• Memory management for LLM agents — how agents store, update, and retrieve information, for test-time adaptation (How Memory Management Impacts LLM Agents, ACL 2026) and long-horizon personalization (DynamicMem).

I’m always excited to connect. Questions, ideas, and early-stage thoughts are all welcome. If you have anything you’d like to discuss, feel free to reach out!