no code implementations • 7 May 2024 • Paul M. Riechers
A technical note aiming to offer deeper intuition for the LayerNorm function common in deep neural networks.
no code implementations • 6 Oct 2023 • Paul M. Riechers
We address the fundamental limits of learning unknown parameters of any stochastic process from time-series data, and discover exact closed-form expressions for how optimal inference scales with observation length.
no code implementations • 25 Mar 2023 • Sarah E. Marzen, Paul M. Riechers, James P. Crutchfield
One conclusion is that large probabilistic state machines -- specifically, large $\epsilon$-machines -- are key to generating challenging and structurally-unbiased stimuli for ground-truthing recurrent neural network architectures.
no code implementations • 10 Aug 2018 • Paul M. Riechers
Framing computation as the transformation of metastable memories, we explore its fundamental thermodynamic limits.
Statistical Mechanics Applied Physics 80-02, 82C03, 82C05, 68Q10 F.0