| Transformers learn to implement preconditioned gradient descent for in-context learning K Ahn, X Cheng, H Daneshmand, S Sra Advances in Neural Information Processing Systems (NeurIPS) 36, 2024 | 172 | 2024 |
| Optimal dimension dependence of the metropolis-adjusted langevin algorithm S Chewi, C Lu, K Ahn, X Cheng, T Le Gouic, P Rigollet Conference on Learning Theory (COLT), 1260-1300, 2021 | 87 | 2021 |
| Understanding the unstable convergence of gradient descent K Ahn, J Zhang, S Sra International Conference on Machine Learning, 247-257, 2022 | 81 | 2022 |
| From Nesterov's Estimate Sequence to Riemannian Acceleration K Ahn, S Sra Proceedings of Thirty Third Conference on Learning Theory (COLT), PMLR 125 …, 2020 | 80 | 2020 |
| Sgd with shuffling: optimal rates without component convexity and large epoch requirements K Ahn, C Yun, S Sra Advances in Neural Information Processing Systems 33, 17526-17535, 2020 | 80 | 2020 |
| Hypergraph spectral clustering in the weighted stochastic block model K Ahn, K Lee, C Suh IEEE Journal of Selected Topics in Signal Processing 12 (5), 959-974, 2018 | 75 | 2018 |
| Efficient constrained sampling via the mirror-Langevin algorithm K Ahn, S Chewi Advances in Neural Information Processing Systems 34, 28405-28418, 2021 | 73 | 2021 |
| Learning threshold neurons via edge of stability K Ahn, S Bubeck, S Chewi, YT Lee, F Suarez, Y Zhang Advances in Neural Information Processing Systems (NeurIPS) 36, 2024 | 44 | 2024 |
| Linear attention is (maybe) all you need (to understand transformer optimization) K Ahn, X Cheng, M Song, C Yun, A Jadbabaie, S Sra International Conference on Learning Representations (ICLR), 2023 | 43 | 2023 |
| Community recovery in hypergraphs K Ahn, K Lee, C Suh IEEE Transactions on Information Theory 65 (10), 6561-6579, 2019 | 43 | 2019 |
| Graph Matrices: Norm Bounds and Applications K Ahn, D Medarametla, A Potechin arXiv preprint 1604.03423, 2020 | 36* | 2020 |
| Binary rating estimation with graph side information K Ahn, K Lee, H Cha, C Suh Advances in neural information processing systems 31, 2018 | 36 | 2018 |
| Mirror descent maximizes generalized margin and can be implemented efficiently H Sun, K Ahn, C Thrampoulidis, N Azizan Advances in Neural Information Processing Systems 35, 31089-31101, 2022 | 24 | 2022 |
| Reproducibility in optimization: Theoretical framework and limits K Ahn, P Jain, Z Ji, S Kale, P Netrapalli, GI Shamir Advances in Neural Information Processing Systems 35, 18022-18033, 2022 | 23 | 2022 |
| Riemannian perspective on matrix factorization K Ahn, F Suarez arXiv preprint arXiv:2102.00937, 2021 | 17 | 2021 |
| Understanding Nesterov's Acceleration via Proximal Point Method K Ahn, S Sra Symposium on Simplicity in Algorithms (SOSA), 117-130, 2022 | 16 | 2022 |
| The crucial role of normalization in sharpness-aware minimization Y Dai, K Ahn, S Sra Advances in Neural Information Processing Systems (NeurIPS) 36, 2024 | 14 | 2024 |
| One-pass learning via bridging orthogonal gradient descent and recursive least-squares Y Min, K Ahn, N Azizan 2022 IEEE 61st Conference on Decision and Control (CDC), 4720-4725, 2022 | 14 | 2022 |
| A Unified Approach to Controlling Implicit Regularization via Mirror Descent H Sun, K Gatmiry, K Ahn, N Azizan arXiv preprint arXiv:2306.13853, 2023 | 12 | 2023 |
| Understanding Adam optimizer via online learning of updates: Adam is FTRL in disguise K Ahn, Z Zhang, Y Kook, Y Dai International Conference on Machine Learning (ICML), 2024 | 10 | 2024 |
Suvrit SraTUM and MITAdresse e-mail validée de mit.edu