Yang Lin
AI / Data Engineer
Building intelligent systems from models to applications.
Selected Work
Projects across language models, generative models, planning and computer vision.
GPT Story Generation
Built a GPT-style language model from scratch with nanoGPT, trained on the ROCStories dataset for 5-sentence story generation.
Flow Matching & MeanFlow
Comparing x-, ε- and v-prediction objectives, and one-step sampling behaviour in high dimensions.
Multi-Drone Motion Planning
Bidirectional RRT-Connect in a 3K-dimensional joint configuration space, evaluated across obstacle scenarios.
Self-Supervised Semantic Segmentation
CNN patch representations, spatial coordinates and K-means clustering — no labels, no pretrained weights.
About
I am a Master's student in Computing at the Australian National University, specialising in Artificial Intelligence and Machine Learning. I expect to graduate in December 2026.
With a background in applied data analytics and statistics, I enjoy turning raw data and research ideas into machine learning models and practical AI systems.
My current interests include large language models, generative models, AI agents, and planning, with a focus on building and experimenting with practical AI applications.
Experience
Analysed user behaviour and engagement data to identify patterns in user activity, retention, and interactions.
Analysed market and competitor data to identify user segments and engagement drivers, providing data-driven recommendations.
Developed research databases and Python/SQL/Azure ETL workflows, reducing manual processing by approximately 75% while maintaining 99%+ data accuracy.