Youngsik Hwang

2 accepted papers

2026

Flatness-Aware Stochastic Gradient Langevin Dynamics

Stefano Bruno, Youngsik Hwang, JaeHyeon An, Sotirios Sabanis, Dongyoung Lim

ICML 2026poster

Flatness of the loss landscape has been widely studied as an important perspective for understanding the behavior and generalization of deep learning algorithms. Motivated by this view, we propose Flatness-Aware Stochastic Gradient Langevin Dynamics (fSGLD), a first-order optimization method that bi…

Cited by 0 Source ↗Scholar ↗

2024

Dual Cone Gradient Descent for Training Physics-Informed Neural Networks

Youngsik Hwang, Dongyoung Lim

NeurIPS 2024poster

Physics-informed neural networks (PINNs) have emerged as a prominent approach for solving partial differential equations (PDEs) by minimizing a combined loss function that incorporates both boundary loss and PDE residual loss. Despite their remarkable empirical performance in various scientific comp…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗