Publications
* indicates equal contribution. See also my Google Scholar.
MuonBP: Faster Muon via Block-Periodic Orthogonalization
A. Khaled, K. Ozkara, T. Yu, Y. Park.
Published in International Conference on Learning Representations (ICLR), 2026.
SPIRE: Conditional Personalization for Federated Diffusion Generative Models
K. Ozkara, R. Zhou, S. Diggavi.
Published in International Conference on Artificial Intelligence and Statistics (AISTATS), 2026.
Directional Alignment Mitigates Reward Hacking in Reinforcement Learning for Language Models
W. Deng, J. Huang, K. Ozkara, Y. Li, C. Thrampoulidis, X. Li, Y. Park.
Published in ICML 2026 Second Workshop on Agents in the Wild: Safety, Security, and Beyond, 2026.
Stochastic Rounding for LLM Training: Theory and Practice
K. Ozkara, T. Yu, Y. Park.
Published in International Conference on Artificial Intelligence and Statistics (AISTATS), 2025.
ADEPT: Hierarchical Bayes Approach to Personalized Federated Unsupervised Learning
K. Ozkara, B. Huang, R. Zhou, S. Diggavi.
Published in International Conference on Artificial Intelligence and Statistics (AISTATS), 2025.
MADA: Meta-Adaptive Optimizers through Hyper-gradient Descent
K. Ozkara, C. Karakus, P. Raman, M. Hong, S. Sabach, B. Kveton, V. Cevher.
Published in International Conference on Machine Learning (ICML), 2024.
A Statistical Framework for Personalized Federated Learning and Estimation: Theory, Algorithms, and Privacy
K. Ozkara*, A. Girgis*, D. Data, S. Diggavi.
Published in International Conference on Learning Representations (ICLR), 2023.
Personalized PCA for Federated Heterogeneous Data
K. Ozkara, B. Huang, S. Diggavi.
Published in IEEE International Symposium on Information Theory (ISIT), 2023.
QuPeD: Quantized Personalization via Distillation with Applications to Federated Learning
K. Ozkara, N. Singh, D. Data, S. Diggavi.
Published in Advances in Neural Information Processing Systems (NeurIPS), 2021.