Howdy! I’m Weijia Zhang

Research

My research studies reliable self-improving agents for long-horizon interactive tasks. I explore how agents can learn from interaction traces and failures, acquire reusable memory and skills, and improve through inference-time evolution and SFT/RL across coding, computer-use, and multimodal environments.

• Self-Evolving Agents: recursive self-improvement (RSI), self-evolving harnesses, memory and skills, and multi-agent evolution
• Multimodal Agents: game-development agents, GUI and computer-use agents, and agentic world models
• Agent Post-Training: SFT and RL for coding agents and long-horizon agents

News

• 2026.08: 🎉 My first-author paper CUADebug was accepted to EMNLP 2026 Findings! 📄✨
• 2026.08: 🎉 AgentDebugX was accepted to EMNLP 2026 (System Demonstration)! 🛠️🤖
• 2026.06: 🎉 Joined TikTok as a Machine Learning Engineer Intern working on self-evolving agents! 🤖✨
• 2026.05: 🎉🎓 Graduated from UIUC in Math + Computer Science with the 2026 C.W. Gear Outstanding Undergraduate Student award ! 🏆
• 2026.01: 🎉👑 Paper on bias inheritance was accepted to ACL 2026 as an Oral! 📄✨
• 2025.07: 🎉 Joined Microsoft Research Asia as a research intern! 🔬💼

Publications

• 

  CUADebug: Diagnosing and Repairing Computer-Use Agent Failures

  Weijia Zhang, Kunlun Zhu, Zeyi Liu, Yinting Chen, Tianyi Ma, Jiateng Liu, Jiaxun Zhang, Bingxuan Li, Pan Lu, Xiangru Tang, Heng Ji, and Jiaxuan You

  EMNLP 2026

  Paper Details

• 

  Understanding and Mitigating the Bias Inheritance in LLM-based Data Augmentation on Downstream Tasks

  Miaomiao Li, Hao Chen, Yang Wang, Tingyuan Zhu, Weijia Zhang, Kaijie Zhu, Kam-Fai Wong, and Jindong Wang

  ACL 2026, Oral

  Paper Details

• 

  AgentDebugX: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in LLM Agents

  Kunlun Zhu, Xuyan Ye, Zhiguang Han, Yuchen Zhao, Bingxuan Li, Weijia Zhang, Muxin Tian, Xiangru Tang, Pan Lu, James Zou, Jiaxuan You, and Heng Ji

  EMNLP 2026 (System Demonstration)

  Paper Code Website Details

• 

  How Much Vision Does Multimodal Reasoning Need? Vision-Stripping for Multimodal Benchmarks

  Weijia Zhang, Zijia Liu, Tianyi Zhang, Ruiqi Chen, Lian Zhang, Haoru Li, Haoqi Chen, and Jiaxuan You

  Under review at NeurIPS 2026

  Details

• 

  Every Act Has Its Price: Compressed Moral Composition in Frontier LLMs

  Weijia Zhang*, Ruiqi Chen*, Yunze Xiao*, and Weihao Xuan

  arXiv:2606.11232

  Paper Details

• 

  Where LLM Agents Fail and How They Can Learn From Failures

  Kunlun Zhu, Zijia Liu, Bingxuan Li, Muxin Tian, Yingxuan Yang, Jiaxun Zhang, Pengrui Han, Qipeng Xie, Fuyang Cui, Weijia Zhang, Xiaoteng Ma, Xiaodong Yu, Gowtham Ramesh, Jialian Wu, Zicheng Liu, Pan Lu, James Zou, and Jiaxuan You

  Failure Modes in Agentic AI Workshop, ICML 2026

  Paper Details

• 

  SeeingEye: Agentic Information Flow Unlocks Multimodal Reasoning in Text-only LLMs

  Weijia Zhang*, Zijia Liu*, Haoru Li*, Haoqi Chen*, and Jiaxuan You

  arXiv:2510.25092 · AAAI 2026 under review

  Paper Details

• 

  Artificial Leviathan: Exploring Social Evolution of LLM Agents Through the Lens of Hobbesian Social Contract Theory

  Gordon Dai*, Weijia Zhang*, Jinhan Li, Siqi Yang, Srihas Rao, Arthur Caetano, and Misha Sra

  The First Workshop on AI Behavioral Science, ACM SIGKDD 2024

  Paper Details

* Equal contribution. See the publications page or Google Scholar for more.

Gamedev

Beyond research, I am a passionate indie game developer. Feel free to check out my work on the game page. I am also happy to discuss the future of AI and games.