Undergraduate Student · Tsinghua University
Building language systems that reason, plan, and act with people.
I'm Zixuan (Alex) Wang, a Mathematics and Physics undergraduate at
Tsinghua University,
with a minor in Artificial Intelligence.
I work on agentic reinforcement learning, agent evaluation, and LLM personalization.
I am applying for 2027 Fall graduate program in the U.S.
I am Zixuan (Alex) Wang, a Mathematics and Physics undergraduate at Tsinghua University (entered in 2023),
with a minor in Artificial Intelligence.
I study agentic reinforcement learning and LLM personalization,
with an interest in how agents learn from interaction and adapt to new tasks and users.
I have worked as a research assistant at 
Carnegie Mellon University, advised by
Prof. Andrea Zanette.
During my Fall 2025 exchange at 
- Mind2Dialogue was accepted to the NeurIPS 2026 UserSim Workshop.
- Harness Learning was accepted to the COLM 2026 LLA Workshop as a Spotlight Oral.
My recent work explores how agents can adapt their harnesses from execution feedback, how user simulation can provide supervision for human-aware models, and how to evaluate the user information retained in long-term agent memory.
Agent Learning
Reinforcement learning for harness adaptation, tool use, and sustained interaction.
Human-Aware Models
User simulation and structured supervision for personalization and reasoning about user intent.
Agent Evaluation
Evaluating adaptation across tasks and the user information agents retain in memory.
Tsinghua UniversityUC San Diego
Papers & Preprints
LLA Workshop
Harness Learning Enables Generalizable Test-Time Adaptation
UserSim Workshop
Mind2Dialogue: Training Human-Aware Language Models by Simulating User Mental States
Preprint
Drift Calibration in Geometric Eye Tracking Systems
Preprint
MemAudit: Auditing Long-Term Agent Memory via Hidden User-State Recovery
Technical Reports
Technical Report
Selected Writing
Selected Writing
RC3: Rollout Chunking with Context Compression for Accelerating Long-Horizon Reinforcement Learning
Technical Blogs
Course Notes
3D Visual Computing Course Notes
Notes on 3D visual computing, including geometry processing, rendering, and 3D reconstruction techniques.
Machine Learning Course Notes - Learning Theory
Comprehensive notes on learning theory, covering PAC learning, VC dimension, and statistical learning foundations.
Course
An Optimization View of DP LLM Fine-tuning: When Does Bias Correction Help, and Can the Optimizer Be Improved?
Course
Ego-embodied Reasoner: Egocentric Embodied Reasoning and Planning with MLLM via Reinforcement Learning
Course
EgoHOI: Prior-Guided 3D Hand-Object Interaction Reconstruction from Monocular Egocentric RGB Video
Course
Unconditional and Image-conditioned 3D Generation
Course
Human Skeleton and Skin Generation
Course
I build with agents, and agents run on tokens. Every chart below is live telemetry from my own machines — every token my coding agents have read, cached, and written.