Daniel Kunin

According to our database, Daniel Kunin authored at least 10 papers between 2019 and 2024.

Collaborative distances:

• Dijkstra number of five.
• Erdős number of four.

2024

The Limiting Dynamics of SGD: Modified Loss, Phase-Space Oscillations, and Anomalous Diffusion.

Daniel Kunin, Javier Sagastuy-Breña, Lauren Gillespie, Eshed Margalit, Hidenori Tanaka, Surya Ganguli, Daniel L. K. Yamins

Neural Comput., January, 2024

2023

Stochastic Collapse: How Gradient Noise Attracts SGD Dynamics Towards Simpler Subnetworks.

Feng Chen, Daniel Kunin, Atsushi Yamamura, Surya Ganguli

Proceedings of the Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, 2023

The Asymmetric Maximum Margin Bias of Quasi-Homogeneous Neural Networks.

Daniel Kunin, Atsushi Yamamura, Chao Ma, Surya Ganguli

Proceedings of the Eleventh International Conference on Learning Representations, 2023

2021

Rethinking the limiting dynamics of SGD: modified loss, phase space oscillations, and anomalous diffusion.

Daniel Kunin, Javier Sagastuy-Breña, Lauren Gillespie, Eshed Margalit, Hidenori Tanaka, Surya Ganguli, Daniel L. K. Yamins

CoRR, 2021

Noether's Learning Dynamics: The Role of Kinetic Symmetry Breaking in Deep Learning.

Hidenori Tanaka, Daniel Kunin

CoRR, 2021

Noether's Learning Dynamics: Role of Symmetry Breaking in Neural Networks.

Hidenori Tanaka, Daniel Kunin

Proceedings of the Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, 2021

Neural Mechanics: Symmetry and Broken Conservation Laws in Deep Learning Dynamics.

Daniel Kunin, Javier Sagastuy-Breña, Surya Ganguli, Daniel L. K. Yamins, Hidenori Tanaka

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

2020

Pruning neural networks without any data by iteratively conserving synaptic flow.

Hidenori Tanaka, Daniel Kunin, Daniel L. K. Yamins, Surya Ganguli

Proceedings of the Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, 2020

Two Routes to Scalable Credit Assignment without Weight Symmetry.

Daniel Kunin, Aran Nayebi, Javier Sagastuy-Breña, Surya Ganguli, Jonathan M. Bloom, Daniel Yamins

Proceedings of the 37th International Conference on Machine Learning, 2020

2019

Loss Landscapes of Regularized Linear Autoencoders.

Daniel Kunin, Jonathan M. Bloom, Aleksandrina Goeva, Cotton Seed

Proceedings of the 36th International Conference on Machine Learning, 2019