Vincent-Daniel Yun

This is Daniel, a first-year MSAI student at the University of Southern California researching Deep Learning Foundation & LLM Optimization. I'm fortunately supervised by Prof. Sai Praneeth Karimireddy and Prof. Vatsal Sharan from USC. I am also closely working with Prof. Sunwoo Lee from Inha University (Republic of Korea).

Recent progress in AI such as large language models has been driven largely by scaling, but this comes with significant computational and memory costs. This motivates a broader question at the core of my research:

Given that modern neural networks are inherently overparameterized and computationally redundant, how can we transform them into structurally and computationally optimal systems?

Therefore, my research focuses on improving optimization, generalization, and model compression techniques to make neural networks lighter, faster, and more efficient in learning from data.

My current research interests:

• Efficient and Robust Multi-Agent
• Efficient LLM Optimization & Reasoning
• Efficient & Effective optimization/training dynamics of neural networks
• Model pruning, quantization, and compression techniques
• Deep learning foundation

My current affiliations include:

• Theory Group at University of Southern California (USC CS Theory Group)
• Large-Scale Machine Learning Systems Lab at Inha University (LMLS)