Summary

M.S. student in Information Science at UIUC, working on ML systems for large-scale models: MoE inference engines and expert streaming, distributed training and rollout infrastructure, RL post-training systems, and GPU compiler and kernel-level correctness. Merged open-source contributions to colibri, DeepSpeed, Triton, verl, and LMCache; research background in LLM reasoning, multimodal learning, and agent safety.

Education

• M.S. in Information Science

  2028-07
  University of Illinois Urbana-Champaign
• B.S. in Business Analytics

  2026-06
  Shanghai University of Finance and Economics

  GPA: 3.71/4.0

Interests

• ML Systems for Large Models

  MoE inference, expert streaming, KV-cache budgeting, performance profiling
• Distributed Training and Rollout

  ZeRO-3, FSDP, Megatron-LM, Hybrid Engine, RL rollout systems
• GPU Compiler and Kernel Correctness

  MLIR, Triton, CUDA
• LLM Reasoning and Multimodal Learning

  chain-of-thought supervision, selective fine-tuning, retrieval and fusion
• Agents and Security

  agentic planning, RAG, execution-grounded evaluation, runtime safety gates