🎓 Graduate student at FCAI Lab, Yonsei University
🔬 Researching AI Model Decomposition & Verification
On the Decompositionality of Neural Networks · arXiv preprint, 2026
Junyong Lee, Baek-Ryun Seong, Sang-Ki Ko, Andrew Ferraiuolo, Minwoo Kang, Hyuntae Jeon, Seungmin Lim, Jieung Kim
Introduces neural decompositionality, a formal semantic-preserving abstraction over neural architectures, and proposes SAVED — a boundary-aware framework for verification-driven decomposition.
Languages & Frameworks
Editor & Agents
