4 papers
Convergence guarantees for Muon: New parameter regimes and generalizations
Arthur C. B. de Oliveira, Dhruv D. Jatkar, Guilherme S. Vicinansa +1
In this paper, we establish the first asymptotic convergence guarantees for the Muon algorithm through a more accurate proxy for the Newton-Schultz iteration than the typical matri…
On the Convergence of Overparameterized Problems: Inherent Properties of the Compositional Structure of Neural Networks
Arthur Castello Branco de Oliveira, Dhruv Jatkar, Eduardo Sontag
This paper investigates how the compositional structure of neural networks shapes their optimization landscape and training dynamics. We analyze the gradient flow associated with o…
Interplay between Contractivity and Monotonicity for Reaction Networks
Alon Duvall, M. Ali Al-Radhawi, Dhruv D. Jatkar +1
We establish a new relationship between monotonicity and contractivity and use this connection to describe a new general class of weakly contractive reaction networks. The new clas…
Competition for binding targets results in paradoxical effects for simultaneous activator and repressor action -- Extended Version
M. Ali Al-Radhawi, Krishna Manoj, Dhruv D. Jatkar +3
In the context of epigenetic transformations in cancer metastasis, a puzzling effect was recently discovered, in which the elimination (knock-out) of an activating regulatory eleme…