KG
Khashayar Gatmiry
cs.LGstat.MLmath.STstat.THcs.DScs.AIcs.CLmath.OCcs.CCcs.GT
On Valency
published · living versionsW_d9p2aje4·v1 · currentpublished
Transformers Learn Two Competing Sorting Algorithms, and Only One Length-Generalizes
with Nathan W. Henry, Chloe H. Su, Emmanuel Abbe, Samy Bengio +3
1 version
Preprints & journals
27 papers in the corpus · 2018–2026Transformers Learn Two Competing Sorting Algorithms, and Only One Length-Generalizes2610.26974v1 · Khashayar Gatmiry, Nathan W. Henry, Chloe H. Su et al.2026 · 0 citationsarXivon Valency
RLTLDR: Self-improvement by Internalizing Self-generated Feedback2609.37633v2 · Michael Kirchhof, Eleonora Gualdoni, Andrew Szot et al.2026 · 0 citationsarXiv
Learn Your Own Thoughts: Abstract Token Curriculum2609.19717v2 · Khashayar Gatmiry, Avrajit Ghosh, Parsa Mirtaheri et al.2026 · 0 citationsarXiv
High-accuracy and dimension-free sampling with diffusions2601.10708v1 · Khashayar Gatmiry, Sitan Chen, Adil Salim2026 · 0 citationsarXiv
Rethinking Invariance in In-context Learning2505.04994v1 · Lizhe Fang, Yifei Wang, Khashayar Gatmiry et al.2025 · 2 citationsarXiv
On the Role of Depth and Looping for In-Context Learning with Task Diversity2410.21698v1 · Khashayar Gatmiry, Nikunj Saunshi, Sashank J. Reddi et al.2024 · 0 citationsarXiv
Computing Optimal Regularizers for Online Linear Optimization2410.17336v1 · Khashayar Gatmiry, Jon Schneider, Stefanie Jegelka2024 · 0 citationsarXiv
Simplicity Bias via Global Convergence of Sharpness Minimization2410.16401v1 · Khashayar Gatmiry, Zhiyuan Li, Sashank J. Reddi et al.2024 · 0 citationsarXiv
Can Looped Transformers Learn to Implement Multi-step Gradient Descent for In-context Learning?2410.08292v1 · Khashayar Gatmiry, Nikunj Saunshi, Sashank J. Reddi et al.2024 · 0 citationsarXiv
What does guidance do? A fine-grained analysis in a simple setting2409.13074v1 · Muthu Chidambaram, Khashayar Gatmiry, Sitan Chen et al.2024 · 1 citationarXiv
Adversarial Online Learning with Temporal Feedback Graphs2407.00571v1 · Khashayar Gatmiry, Jon Schneider2024 · 0 citationsarXiv
Learning Mixtures of Gaussians Using Diffusion Models2404.18869v2 · Khashayar Gatmiry, Jonathan Kelner, Holden Lee2024 · 1 citationarXiv
EM for Mixture of Linear Regression with Clustered Data2308.11518v1 · Amirhossein Reisizadeh, Khashayar Gatmiry, Asuman Ozdaglar2023 · 0 citationsarXiv
A Unified Approach to Controlling Implicit Regularization via Mirror Descent2306.13853v2 · Haoyuan Sun, Khashayar Gatmiry, Kwangjun Ahn et al.2023 · 2 citationsarXiv
The Inductive Bias of Flatness Regularization for Deep Matrix Factorization2306.13239v1 · Khashayar Gatmiry, Zhiyuan Li, Ching-Yao Chuang et al.2023 · 1 citationarXiv
Projection-Free Online Convex Optimization via Efficient Newton Iterations2306.11121v1 · Khashayar Gatmiry, Zakaria Mhammedi2023 · 0 citationsarXiv
When does Metropolized Hamiltonian Monte Carlo provably outperform Metropolis-adjusted Langevin algorithm?2304.04724v3 · Yuansi Chen, Khashayar Gatmiry, Minhui Jiang2023 · 0 citationsarXiv
A Simple Proof of the Mixing of Metropolis-Adjusted Langevin Algorithm under Smoothness and Isoperimetry2304.04095v2 · Yuansi Chen, Khashayar Gatmiry2023 · 0 citationsarXiv
Sampling with Barriers: Faster Mixing via Lewis Weights2303.00480v2 · Khashayar Gatmiry, Jonathan Kelner, Santosh S. Vempala2023 · 0 citationsarXiv
Near-Optimal Algorithms for Group Distributionally Robust Optimization and Beyond2212.13669v2 · Tasuku Soma, Khashayar Gatmiry, Sharut Gupta et al.2022 · 1 citationarXiv
Bandit Algorithms for Prophet Inequality and Pandora's Box2211.08586v2 · Khashayar Gatmiry, Thomas Kesselheim, Sahil Singla et al.2022 · 11 citationsarXiv
Quasi-Newton Steps for Efficient Online Exp-Concave Optimization2211.01357v3 · Zakaria Mhammedi, Khashayar Gatmiry2022 · 0 citationsarXiv
On the generalization of learning algorithms that do not converge2208.07951v2 · Nisha Chandramoorthy, Andreas Loukas, Khashayar Gatmiry et al.2022 · 3 citationsarXiv
Convergence of the Riemannian Langevin Algorithm2204.10818v1 · Khashayar Gatmiry, Santosh S. Vempala2022 · 2 citationsarXiv
Testing Determinantal Point Processes2008.03650v1 · Khashayar Gatmiry, Maryam Aliakbarpour, Stefanie Jegelka2020 · 0 citationsarXiv
Non-submodular Function Maximization subject to a Matroid Constraint, with Applications1811.07863v5 · Khashayar Gatmiry, Manuel Gomez-Rodriguez2018 · 8 citationsarXiv
Information Theoretic Bounds on Optimal Worst-case Error in Binary Mixture Identification1811.07307v3 · Khashayar Gatmiry, Seyed Abolfazl Motahari2018 · 0 citationsarXiv
Career total: 31 works. 27 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.