no code implementations • 27 May 2024 • Jessica N. Howard, Marc S. Klinger, Anindita Maiti, Alexander G. Stapleton
In this paper, we unify NNFT and BRG to form a powerful new framework for exploring the space of NNs and SFTs, which we coin BRG-NNFT.
1 code implementation • 20 May 2024 • Yue M. Lu, Mary I. Letey, Jacob A. Zavatone-Veth, Anindita Maiti, Cengiz Pehlevan
Transformers have a remarkable ability to learn and execute tasks based on examples provided within the input itself, without explicit prior training.
no code implementations • 9 May 2024 • Jessica N. Howard, Ro Jefferson, Anindita Maiti, Zohar Ringel
We systematically integrate out the unlearnable modes of the GP kernel, thereby obtaining an RG flow of the Gaussian Process in which the data plays the role of the energy scale.
no code implementations • 6 Jul 2023 • Mehmet Demirtas, James Halverson, Anindita Maiti, Matthew D. Schwartz, Keegan Stoner
Conversely, the correspondence allows one to engineer architectures realizing a given field theory by representing action deformations as deformations of neural network parameter densities.
1 code implementation • 1 Jun 2021 • Anindita Maiti, Keegan Stoner, James Halverson
We demonstrate that symmetries of network densities may be determined via dual computations of network correlation functions, even when the density is unknown and the network is not equivariant.
1 code implementation • 19 Aug 2020 • James Halverson, Anindita Maiti, Keegan Stoner
We propose a theoretical understanding of neural networks in terms of Wilsonian effective field theory.