cv

B.S.

University of Illinois Urbana-Champaign (UIUC)

Computer Science and Mathematics

• Algorithms
• Systems Programming
• Networks
• Database Systems
• Data Structures
• Computer Architecture
• Probability
• Numerical Analysis
• Programming Language & Compilers
• ML
• NLP

Research Intern in Data, Knowledge, and Intelligence Group

Microsoft

Research on multimodal agent in chart for Microsoft Copilot in Excel.

Software Engineer Intern in Wechat Group

Tencent

Developed a hardware performance tool and Unity Memory performance tool for Wechat Mini Program.

Software Engineer Intern

NetEase

Designed and developed a full-stack game art assets database with dependency graph using SQLite, Flask, and Vue.js.

AI Software Developer Intern

Reborn Network

• Developed multiplayer Role-playing agents framework enabling dynamic interactions (text, voice, action) with AI NPCs in VR environments, including IK and function calling support for physical and directive user engagements. • Designed and implemented seamless voice interactions (<1s latency) in VR, enhancing real-time AI agent responsiveness. • Managed long-term memory using vector database with dynamic ADD/UPDATE/DELETE operations. • Designed character card system for importing/exporting/sharing NPC personas.

OpenManus-RL: Better Memory Planning Agent in General Environments via Reinforcement Learning

UIUC U Lab

Implemented REACT reasoning architecture (think→act→tool-use) and multi-agent flow for action optimization. Built MCP-compatible tool framework supporting automated tool invocation and output processing. Constructed SFT datasets compatible with AgentBench environment (Gaia, WebArena, WebShop, AlfWorld).

Understanding and Mitigating Bias Inheritance in LLM-based Data Augmentation

UIUC U Lab

Executed 30+ fine-tuning experiments using Llama-3.1 and ChatGPT-o1-mini with controlled bias blend ratios. Developed automated pipeline for synthetic data generation, bias injection, LoRA fine-tuning, and evaluation. Implemented and compared three bias mitigation strategies: token-based, mask-based, and loss-based approaches.

Artificial Leviathan: LLM Agent Social Evolution via Hobbesian Contract Theory

UIUC U Lab

Designed multi-agent social simulation framework to model social contract formation. Implemented persona-encoding via prompt engineering and personalized QA sets for differentiated agent personalities. Developed agent economic system and AI policy generation framework.