معرفی
Vegard Antun is a Postdoctoral Fellow at the Department of Mathematics, University of Oslo, specializing in applied mathematics with a focus on inverse problems, imaging, and deep learning.
- Education: PhD (2020), Master's (2016), and Bachelor's (2013) degrees from the University of Oslo.
- Research Interests: Stability and accuracy in AI algorithms, compressive sensing, signal recovery, and mathematical paradoxes in deep learning.
- Key Projects: Supervised a 2022 interdisciplinary project on deep learning observables for partial differential equations.
His work explores the theoretical limitations of AI, particularly the instability of neural networks in inverse problems and their implications for scientific computing, as highlighted in his research on mathematical paradoxes and Smale’s 18th problem.
His publications span topics such as binary sampling, wavelet reconstruction, and data-efficient neural networks, emphasizing the tension between AI accuracy and robustness.
He has contributed to understanding implicit regularization, existence of optimal decoders, and hybrid concept-based models for scientific applications.