E-mail: samet.demir@austin.utexas.edu

Address: POB 6.244, Oden Institute, UT Austin
Austin, TX / USA

I’m a Peter O’Donnell, Jr. Postdoctoral Fellow at the Oden Institute for Computational Engineering & Sciences, The University of Texas at Austin, working with Prof. Blake Bordelon. I received my Ph.D. in Computer Science & Engineering from Koc University in 2026, advised by Prof. Zafer Dogan, where I was an AI Fellow at the KUIS AI Research Center and a senior member of the Machine Learning & Information Processing (MLIP) group. Previously, I received B.S. in Computer Engineering from Boğaziçi University.

My research develops a high-dimensional learning theory for how the statistics of learned representations arise from the interplay of architecture, data, and optimization—and how these statistics govern behaviors like generalization, in-context learning, and robustness in modern deep learning.

Research interests

Learning theory High-dimensional statistics Deep learning In-context learning Gaussian universality Random matrix theory Robust learning

View full CV

Open to research positions starting Fall 2027

Selected publications

• Demir & Dogan. Optimal Attention Temperature Improves the Robustness of In-Context Learning under Distribution Shift in High Dimensions. ICML 2026.
• Demir & Dogan. How Data Mixing Shapes In-Context Learning: Asymptotic Equivalence for Transformers with MLPs. NeurIPS 2025.
• Demir & Dogan. Asymptotic Analysis of Two-Layer Neural Networks after One Gradient Step under Gaussian Mixtures with Structure. ICLR 2025.
• Demir & Dogan. Input-Label Correlation Governs a Linear-to-Nonlinear Transition in Random Features under Spiked Covariance. SIAM Journal on Mathematics of Data Science, 2026.
• Demir & Dogan. Implicitly Normalized Online PCA: A Regularized Algorithm with Exact High-Dimensional Dynamics. arXiv preprint, arXiv:2512.01231, 2025.
• More on Google Scholar.

Awards & fellowships

• Peter O’Donnell, Jr. Postdoctoral Fellowship, Oden Institute, UT Austin (2026–27).
• AI Fellowship, KUIS AI Research Center, Koc University.

Academic service

• Reviewer: ICML 2026, ICLR 2025–26, NeurIPS 2025-26, AAAI 2026, ACML 2024–25.

Teaching

• Head TA: Introduction to Modern Learning Theory (2025–26); Programming for Engineers (2024–26).
• TA: Algorithms & Complexity (2023–24), Software Engineering (2023), Programming with Python (2022), Computer Systems & Programming (2020–22), Introduction to Computing (2017).