Alert button
Picture for Aryan Mokhtari

Aryan Mokhtari

Alert button

In-Context Learning with Transformers: Softmax Attention Adapts to Function Lipschitzness

Feb 18, 2024
Liam Collins, Advait Parulekar, Aryan Mokhtari, Sujay Sanghavi, Sanjay Shakkottai

Viaarxiv icon

An Accelerated Gradient Method for Simple Bilevel Optimization with Convex Lower-level Problem

Feb 12, 2024
Jincheng Cao, Ruichen Jiang, Erfan Yazdandoost Hamedani, Aryan Mokhtari

Viaarxiv icon

Krylov Cubic Regularized Newton: A Subspace Second-Order Method with Dimension-Free Convergence Rate

Jan 05, 2024
Ruichen Jiang, Parameswaran Raman, Shoham Sabach, Aryan Mokhtari, Mingyi Hong, Volkan Cevher

Viaarxiv icon

Projection-Free Methods for Stochastic Simple Bilevel Optimization with Convex Lower-level Problem

Aug 15, 2023
Jincheng Cao, Ruichen Jiang, Nazanin Abolfazli, Erfan Yazdandoost Hamedani, Aryan Mokhtari

Viaarxiv icon

Provable Multi-Task Representation Learning by Two-Layer ReLU Neural Networks

Jul 17, 2023
Liam Collins, Hamed Hassani, Mahdi Soltanolkotabi, Aryan Mokhtari, Sanjay Shakkottai

Viaarxiv icon

Limited-Memory Greedy Quasi-Newton Method with Non-asymptotic Superlinear Convergence Rate

Jun 27, 2023
Zhan Gao, Aryan Mokhtari, Alec Koppel

Figure 1 for Limited-Memory Greedy Quasi-Newton Method with Non-asymptotic Superlinear Convergence Rate
Figure 2 for Limited-Memory Greedy Quasi-Newton Method with Non-asymptotic Superlinear Convergence Rate
Figure 3 for Limited-Memory Greedy Quasi-Newton Method with Non-asymptotic Superlinear Convergence Rate
Figure 4 for Limited-Memory Greedy Quasi-Newton Method with Non-asymptotic Superlinear Convergence Rate
Viaarxiv icon

Accelerated Quasi-Newton Proximal Extragradient: Faster Rate for Smooth Convex Optimization

Jun 03, 2023
Ruichen Jiang, Aryan Mokhtari

Figure 1 for Accelerated Quasi-Newton Proximal Extragradient: Faster Rate for Smooth Convex Optimization
Viaarxiv icon

Greedy Pruning with Group Lasso Provably Generalizes for Matrix Sensing and Neural Networks with Quadratic Activations

Mar 20, 2023
Nived Rajaraman, Devvrit, Aryan Mokhtari, Kannan Ramchandran

Figure 1 for Greedy Pruning with Group Lasso Provably Generalizes for Matrix Sensing and Neural Networks with Quadratic Activations
Figure 2 for Greedy Pruning with Group Lasso Provably Generalizes for Matrix Sensing and Neural Networks with Quadratic Activations
Viaarxiv icon

Online Learning Guided Curvature Approximation: A Quasi-Newton Method with Global Non-Asymptotic Superlinear Convergence

Feb 16, 2023
Ruichen Jiang, Qiujiang Jin, Aryan Mokhtari

Figure 1 for Online Learning Guided Curvature Approximation: A Quasi-Newton Method with Global Non-Asymptotic Superlinear Convergence
Viaarxiv icon