Atish Agarwala

According to our database, Atish Agarwala authored at least 18 papers between 2020 and 2025.

Collaborative distances:

• Dijkstra number of four.
• Erdős number of three.

2025

Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks.

Shikai Qiu, Lechao Xiao, Andrew Gordon Wilson, Jeffrey Pennington, Atish Agarwala

CoRR, July, 2025

Avoiding spurious sharpness minimization broadens applicability of SAM.

Sidak Pal Singh, Hossein Mobahi, Atish Agarwala, Yann N. Dauphin

CoRR, February, 2025

How far away are truly hyperparameter-free learning algorithms?

Priya Kasimbeg, Vincent Roulet, Naman Agarwal, Sourabh Medapati, Fabian Pedregosa, Atish Agarwala, George E. Dahl

Trans. Mach. Learn. Res., 2025

To Clip or not to Clip: the Dynamics of SGD with Gradient Clipping in High-Dimensions.

Noah Marshall, Ke Liang Xiao, Atish Agarwala, Elliot Paquette

Proceedings of the Thirteenth International Conference on Learning Representations, 2025

2024

Feature learning as alignment: a structural property of gradient descent in non-linear neural networks.

Daniel Beaglehole, Ioannis Mitliagkas, Atish Agarwala

Trans. Mach. Learn. Res., 2024

Exact Risk Curves of signSGD in High-Dimensions: Quantifying Preconditioning and Noise-Compression Effects.

Ke Liang Xiao, Noah Marshall, Atish Agarwala, Elliot Paquette

CoRR, 2024

A Clipped Trip: the Dynamics of SGD with Gradient Clipping in High-Dimensions.

Noah Marshall, Ke Liang Xiao, Atish Agarwala, Elliot Paquette

CoRR, 2024

High dimensional analysis reveals conservative sharpening and a stochastic edge of stability.

Atish Agarwala, Jeffrey Pennington

CoRR, 2024

Gradient descent induces alignment between weights and the empirical NTK for deep non-linear networks.

Daniel Beaglehole, Ioannis Mitliagkas, Atish Agarwala

CoRR, 2024

Stepping on the Edge: Curvature Aware Learning Rate Tuners.

Vincent Roulet, Atish Agarwala, Jean-Bastien Grill, Grzegorz Swirszcz, Mathieu Blondel, Fabian Pedregosa

Proceedings of the Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, 2024

Neglected Hessian component explains mysteries in sharpness regularization.

Yann N. Dauphin, Atish Agarwala, Hossein Mobahi

Proceedings of the Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, 2024

2023

Temperature check: theory and practice for training models with softmax-cross-entropy losses.

Atish Agarwala, Samuel Stern Schoenholz, Jeffrey Pennington, Yann N. Dauphin

Trans. Mach. Learn. Res., 2023

On the Interplay Between Stepsize Tuning and Progressive Sharpening.

Vincent Roulet, Atish Agarwala, Fabian Pedregosa

CoRR, 2023

Second-order regression models exhibit progressive sharpening to the edge of stability.

Atish Agarwala, Fabian Pedregosa, Jeffrey Pennington

Proceedings of the International Conference on Machine Learning, 2023

SAM operates far from home: eigenvalue regularization as a dynamical phenomenon.

Atish Agarwala, Yann N. Dauphin

Proceedings of the International Conference on Machine Learning, 2023

2022

Deep equilibrium networks are sensitive to initialization statistics.

Atish Agarwala, Samuel S. Schoenholz

Proceedings of the International Conference on Machine Learning, 2022

2021

One Network Fits All? Modular versus Monolithic Task Formulations in Neural Networks.

Atish Agarwala, Abhimanyu Das, Brendan Juba, Rina Panigrahy, Vatsal Sharan, Xin Wang, Qiuyi Zhang

Proceedings of the 9th International Conference on Learning Representations, 2021

2020

Learning the gravitational force law and other analytic functions.

Atish Agarwala, Abhimanyu Das, Rina Panigrahy, Qiuyi Zhang

CoRR, 2020