작은 질문에서,
분명한 변화로 Small Inquiries,
Substantial Impact.
Small Inquiries, Substantial Impact. 작은 질문에서, 분명한 변화로
안 되는 이유를 찾기보다, 되게 만들어봅니다.Instead of explaining why it can't work, I build it and see.
now AI융합학과AI major ·· 넓은 우주를 탐험하고 있습니다.Exploring a vast universe
Selected workSelected work
- 7xz Lab 2026 A neural network you train, break, and rebuild in the browser, with every forward pass, gradient, and optimizer written by hand. Prune a neuron and it shows the loss change three ways: the direct before/after difference, a path-integral estimate, and a second-order estimate. open the lab →
- Neural network foundational theory 2026 · ongoing Finding an exact, approximation-free relation for the loss change under an arbitrary weight perturbation, and validating it on pruning and quantization. research
- Context-Aware Zone Alert 2025 Alerts on a zone intrusion only when a person is present alongside the target, cutting the false alarms that animals and the environment cause. shipped
최근 기록 Recent log