Li Zizhen

National University of Singapore

Aug 2022 - Aug 2026, Expected

Bachelor of Computing in Computer Science (AI Track)

• CS Courses: AI, Machine Learning, Computer Vision, NLP, Data Structures & Algorithms, etc.
• Math Courses: Discrete Structures, Linear Algebra, Probability, Statistics, Calculus, etc.
• Summer Exchange: Peking University (Machine Learning Algorithms: From Math to Code)
• Standardized Exams: GRE 330/340, TOEFL 109/120 (S:25)
• AY24/25 S2 Dean’s List (Top 5%)
• Certificate of Distinction in AI Focus Area
• NUS STEER Award (S$1000)

TECHNICAL SKILLS

• Languages: Java, Python, C, MATLAB, R
• Frameworks/Libraries: PyTorch, Flask
• Other Tools: Git, GitHub, Bash, Web Crawlers, HTML, CSS

FYP — Retrieval Augmented Generation (RAG)

Aug 2025 - Present · NUS

• Reviewed recent RAG & RL literature to guide system design
• Built end-to-end pipeline: submodular retrieval selector → evidence compression → structured claim generation → format enforcement
• Submitted multiple system runs to TREC AG 2025
• Exploring RL fine-tuning of LLMs in RAG with PPO + LoRA

Cornell — Seeker: Efficient Knowledge Extraction & Retrieval

Jun 2025 - Present · Research Assistant

• Modernized METAM into a modular, reproducible codebase; fixed core bugs and implemented missing features
• Shipped a researcher-facing web UI (end-to-end pipeline, dataset upload, variant toggles, cached re-runs)
• Designed and ran comparative studies across querying, profiling, and weighting variants with analyses/visualizations
• Funded by Cornell IRIP Scholarship ($11,040)

NUS UROP: Discovering Winning Strategies in Sports

Apr 2024 - Apr 2025

• Automated end-to-end pipeline for daily processing, visualization, and reporting of tennis match data (Python, Crontab, Bash)
• Deployed a Flask web platform for accessible player reports
• Built an LLM-powered chatbot for on-demand quantitative analysis and tactical recommendations
• Presented a poster at UREC 2025 and delivered an oral presentation at ISACE 2025
• Website Link | Paper Link

Tsinghua–Berkeley: Active Vision & Optical Zoom

Dec 2024 - Jan 2025

• Trained YOLO for object detection; developed an adaptive zoom algorithm with exponential smoothing
• Conducted feasibility study with the RoboTwin platform
• Improved task success rates by 14% using zoom-augmented data (DP3 baseline)
• Extended AV-ALOHA for VR teleoperation (Quest 3 + USB optical-zoom camera)
• Work submitted to ICRA;
• Paper Link

Tsinghua–Berkeley: Humanoid Robots & Imitation Learning

Jul 2024 - Aug 2025

• Reproduced “OPEN TEACH” for versatile teleoperation
• Integrated Nvidia Isaac Lab/Gym with Meta Quest 3 for gesture-based control of a virtual Allegro hand
• Collected multimodal data (end-effector trajectories, joint angles) for downstream imitation learning
• Applied behavioral cloning and inverse reinforcement learning to improve manipulation performance

X-Institute Summer Camp: Drone Simulation

Jul 2023 - Aug 2023

• Engineered a simulation game with controllable drones exploring 3D terrains
• Developed a vision-based obstacle avoidance system
• Investigated 3D terrain reconstruction for improved navigation
• Presented research as an outstanding representative
• Awarded the 01 Rising Star Award

AI Algorithm Intern, Bytedance

Dec 2023 - Jan 2024

• Developed a web crawler to build a custom handwritten digits dataset
• Trained CNN for digit classification using PyTorch (VGG16, ResNet, Inception v3)
• Created a simulated environment for drone route optimization via DQN
• Evaluated multiple object detection and human pose estimation models

EduLink-NUS Academic Data Management

Jan 2024 - Aug 2024

• Developed a desktop app to streamline student data management for NUS instructors
• Implemented tagging functionalities for efficient data retrieval
• Enhanced UI using JavaFX; performed robust unit & integration testing
• Maintained documentation for developers and end-users
• Project link

Mathematical Contest in Modeling

Feb 2024

• Developed a momentum quantification model for tennis matches (Exponential Decay + AHP)
• Achieved F1 score of 0.75 using a Random Forest-LSTM hybrid model

RightShip Hackathon

Jan 2024

• Analyzed GHG emissions reduction via maritime port electrification with Python
• Certificate

NUS Student Union commIT Technical Team

Dec 2024 - Present

• Developing NUS Student Union website and finance manager software for clubs

NUS Chinese Society

Jan 2024 - Present

• Managed event logistics, transporting goods and arranging venues

Achievements

• National Gold (Division 2) in PHYSICSBOWL 2021
• Gold in British Physics Olympiad 2020-2021 Senior Physics Challenge
• High Distinction in Australian Science Olympiads-Physics
• Entrepreneurship Award at FRC; Motivate Award & Second Place Alliance at FTC
• Shenzhen Foreign Languages School Soccer Team & Club Leader