Curriculum Vitae
Download PDFProfile · Education · Research · Publications · Projects · Skills · Honors · Talks · Teaching · Service
Research Profile
I build reasoning systems that integrate symbolic structure with neural learning. My current work centers on autonomous agents, reinforcement learning and reward modeling, ML mechanics for differentiable logic, formal methods for reliable AI, and quantum computing as a secondary direction.
Education
Stony Brook University, Stony Brook, NY, USA
Ph.D. in Computer Engineering
Aug 2026 - Present
Nanjing University, Nanjing, China
B.Sc. in Information and Computing Sciences (Computer Science Track)
School of Computer Science
Sept. 2022 - June 2026
Selected through China's Pilot Reform Program of Enrollment for Fundamental Disciplines (强基计划), an elite admissions track for foundational sciences. The Computer Science track of the Information and Computing Sciences program is administered by NJU's School of Computer Science and follows the standard Computer Science curriculum.
- GPA 4.60/5.00; Rank 3/25; TOEFL 101 (R29/L26/S23/W23)
- Selected coursework: Machine Learning (95), Principles & Applications of LLMs (92), Knowledge Representation (96), Formal Languages & Automata (94), Operating Systems (91)
Research Experience
Microsoft Research Asia (MSRA)
Research Intern
Advisor: Ziyu Zhou · Apr 2026 - Present
- Working on an AI4Math mathematical research collaboration platform
- Exploring AI systems that support mathematical reasoning and research collaboration
Ludwig Maximilian University of Munich
Research Intern
Advisor: Dr. Yao Zhang · Nov 2025 - Present
- Conducting research on reasoning-based WebAgents with fine-grained reinforcement learning strategies
- Designing reward modeling mechanisms for RL-based WebAgent systems and vertical agent applications
- Studying image-grounded reasoning, multimodal integration, mechanistic analysis of agent harnesses, and formal-methods-inspired world models for reliable agents
KRistal Group, Nanjing University
Research Intern
Advisor: Prof. YiZheng Zhao · Oct 2025 - Present
- Led an empirical and theoretical study of fuzzy operator selection in differentiable logic for neuro-symbolic learning
- Analyzed how implication operators affect gradient routing, optimization behavior, and learning outcomes under different structure / objective / grounding regimes
- First-author manuscript completed; under submission
Independent Research
Research Intern
Advisor: Dr. Zhen Han · Jul 2025 - Present
- Researching reasoning-enhanced reward models for LLMs
- Designed pipelines combining reject sampling, SFT, and RL for scalable preference alignment
- Leading experiments on reasoning-guided reward modeling (manuscript in preparation)
ScaleML Lab, University of Illinois Urbana-Champaign
Research Intern
Advisor: Prof. Tong Zhang · Apr 2025 - Jun 2025
- Built a prototype integrating Lean4/mathlib with LLMs for interactive theorem proving on MiniF2F
- Designed an interactive workflow with state-based verification and closed-loop refinement
- Implemented a bidirectional communication pipeline (LLM ↔ Lean4)
Huawei Central Software Institute, 2012 Labs
Research Intern
Supervisor: JianFeng Gui · Jul 2025 - Sept 2025
- Co-developed a benchmarking system to evaluate reasoning and adaptability of HarmonyOS intelligent agents
- Contributed to the IntelliOS-agent pipeline and ported Python dependency libraries to HarmonyOS
- Integrated HDC debugging pipelines with LLM-based reasoning modules
QUEST Lab, NC State University
Research Intern
Advisor: Prof. Yuan Liu · Jul 2024 - Nov 2024
- Explored quantum memory architectures for quantum machine learning
- Proposed optimized computational architecture for ML workloads
- Co-authored a manuscript on quantum memory architectures (work later continued by collaborators)
COSEC Research Group, Nanjing University
Research Assistant
Advisors: Prof. Yuan Zhang & Prof. Sheng Zhong · Jul 2023 - Dec 2024
- Researched adversarial backdoors in machine learning models
- Proposed an exploit mechanism and designed attack experiments for malicious training pipelines, contributing to the group's broader work on ML robustness and trustworthiness
Publications & Preprints
QuoteBench: How Matched Scores Can Hide Command-Path Failures
Shangao Li, Yao Zhang, Volker Tresp, Yuanyuan Yang
arXiv preprint, 2026 · arXiv:2608.13547
Implication Is Architecture, Not Semantics: Gradient Routing in Differentiable Logic
Shang'ao Li, Yizheng Zhao
Manuscript completed; under submission, 2026
Quantum memory architectures for machine learning workloads
Co-first-author manuscript submitted to ISCA 2025; work later continued by collaborators
Selected Projects
QuoteBench: Benchmarking Shell Quoting in LLM Agents · Jul 2026
Python · Bash · execution-based evaluation
A 56-task execution-verified benchmark separating raw shell competence from failures introduced by JSON and wrapped tool harnesses, across three transport contracts. Released as an arXiv preprint with rollouts on HuggingFace.
CaveBridge: LLM Interface for Colossal Cave Adventure · Jun 2026
Python · LLM intent parsing / narration · Open Adventure deterministic engine
A natural-language front-end for the 1977 Colossal Cave Adventure: the LLM translates free-form Chinese/English input into canonical actions and narrates engine output, while the deterministic kernel stays the source of truth.
LangMatch: Classical Chinese as an LLM Compression Prompt · May 2026
Python · uv · OpenAI API · Transformers (Qwen)
An empirical study of whether a denser prompt language (Classical Chinese) yields a better quality–cost frontier under fixed tasks and model backends; found it a conditional, model-dependent compression strategy.
VocoType-linux: Offline Chinese Speech Input Method for Linux · Dec 2025 – Present
Python · PyTorch · ONNX Runtime · ALSA · IBus / Fcitx5
A high-performance, fully offline Chinese speech input method for Linux (audio capture, VAD, ASR, post-processing, visual config); runs locally with no cloud dependency and has earned 200+ GitHub stars.
Gaussian SPH Fluid: Physics-integrated 3D Gaussians · Winter 2025
C++ · CUDA · OpenGL · 3D Gaussian Splatting · DFSPH
Co-designed a simulation-rendering pipeline that advances 3D Gaussians with a divergence-free SPH solver and renders them directly, with uniform internal filling and SPH-gradient-based covariance updates.
Optimizing Transfer Learning for Healthcare AI Under Data Scarcity · Summer 2025
PyTorch · ViT-B/16 · Masked Autoencoder · AdamW · AMP
Built a skin-lesion classifier on the ISIC dermatoscopic dataset with a self-supervised MAE backbone, two-stage fine-tuning, class-imbalance handling, and post-hoc decision calibration.
Mobile Robot Pose Estimation · Mar – Apr 2025
Python · PCL · ROS
Implemented a point-cloud localization module for robotic manipulation, including statistical outlier removal, voxel downsampling, and ICP-based pose refinement.
Systems and ML Foundations Projects
C++17 · Linux kernel labs · Vue / Spring Boot · LLaMA · vector databases
Implemented a PyTorch-style reverse-mode autograd engine and NN library in C++; completed NJUOS / MIT xv6 scheduling, memory, and file-system labs; built a full-stack retrieval-augmented Q&A / summarization system.
Skills
- Programming: C/C++, Python (primary); Rust, Java, C#, Lean4, Lua
- ML/AI: PyTorch, HuggingFace, FSDP, DeepSpeed, Lean4/mathlib
- Systems: Linux (Arch, Fedora), HPC cluster & distributed training environments
- Languages: Chinese (native), English (fluent, TOEFL 101), Japanese (intermediate, N4, conversational and reading), German (beginner, conversational)
Honors & Awards
- National Scholarship, Ministry of Education of China (2025) - Top 0.2% of students
- First Prize (National Champion), HITCTF 2025 (Antiy Cup) - Ranked 1st nationwide (2025)
- National First Prize, Entropy Cup (熵密杯) Cryptography Challenge, Chinese Association for Cryptologic Research (2025)
- Outstanding Student Pacesetter (university-level), Nanjing University (2024-2025)
- National Third Prize (Finals), CISCN National College Student Information Security Competition (2025)
- First Prize (Semifinal), National Software System Security Challenge (2025)
- National Third Prize, China Graduate Cybersecurity Innovation Competition (2024)
- Outstanding Student (university-level), Nanjing University (2023-2024)
- Special Scholarship for Fundamental Subjects, First Prize (twice) (2023-2025)
- People's Scholarship (twice) (2023-2025)
- Provincial Award, China College Student Computer Design Competition (2023)
- Outstanding Student (college-level), Freshman College, Nanjing University (2022-2023)
- Outstanding Member, Kaijia College Student Union, Nanjing University (2022-2023)
Invited Talks
- Reinforcement Learning with GRPO: From PPO to Group-Relative Policy Optimization, NJU AIA, 2026
- Building a Neural Network from Scratch with NumPy, NJU AIA, 2025
- Building a Neural Network from Scratch with NumPy, NJU AIA, 2023
Teaching Experience
Guest Lectures:
- Lean4 for Interactive Theorem Proving · Discrete Mathematics, Nanjing University · January 2026
- Network Security & Offensive/Defensive Techniques, Nanjing University · December 2025
Teaching Assistant:
- Introduction to Artificial Intelligence, Nanjing University (Senior Lecturer Ye Tao) · Spring 2026
- Cryptography, Nanjing University (Prof. Yuan Zhang) · 2024-2025
- Introduction to Information and Computing Science, Nanjing University (Prof. Sheng Zhong) · 2023-2024
- Discrete Mathematics, Nanjing University (Prof. Sheng Zhong) · 2023-2024
Assisted in homework and exam grading, exam preparation, held weekly exercise classes, and provided student support.
Leadership & Service
- Founder & President, NJU Artificial Intelligence Association (AIA) (2023-2025) - established in 2024 by merging the former campus AI association, with the JSAI NJU Branch as its predecessor
- Co-founder, Jiangsu Society of Artificial Intelligence (JSAI), NJU Branch (2023)
- Chair, School of Computer Science Student Union, Nanjing University (2024-2025)
- Chair, Kaijia College Student Union, Nanjing University (2023-2024)
- Head of Science & Technology Innovation Division, Department of Computer Science Student Union, Nanjing University (2023-2024)
- Captain, Trinity CTF Team, Nanjing University (2024-2026)
- President, NJU Phantom Magic Association (2023-2024)
- Academic Commissary, Information and Computing Sciences Program, Nanjing University (2022-2026)