Search Results for author: Ido Cohen

Found 8 papers, 1 papers with code

Enhancing Neural Training via a Correlated Dynamics Model

no code implementations20 Dec 2023 Jonathan Brokman, Roy Betser, Rotem Turjeman, Tom Berkov, Ido Cohen, Guy Gilboa

Our modeling can improve training efficiency and lower communication overhead, as shown by our preliminary experiments in the context of federated learning.

Federated Learning Image Classification

The Underlying Correlated Dynamics in Neural Training

no code implementations18 Dec 2022 Rotem Turjeman, Tom Berkov, Ido Cohen, Guy Gilboa

We propose in this work a model based on the correlation of the parameters' dynamics, which dramatically reduces the dimensionality.

Dimensionality Reduction

BASiS: Batch Aligned Spectral Embedding Space

no code implementations CVPR 2023 Or Streicher, Ido Cohen, Guy Gilboa

We analyze the degrees of freedom of learning this task using batches and propose a stable alignment mechanism that can work both with batch changes and with graph-metric changes.

Understanding Transformer Memorization Recall Through Idioms

1 code implementation7 Oct 2022 Adi Haviv, Ido Cohen, Jacob Gidron, Roei Schuster, Yoav Goldberg, Mor Geva

In this work, we offer the first methodological framework for probing and characterizing recall of memorized sequences in transformer LMs.

Memorization

Unsupervised Detection of Sub-Territories of the Subthalamic Nucleus During DBS Surgery with Manifold Learning

no code implementations23 Aug 2022 Ido Cohen, Dan Valsky, Ronen Talmon

Compared to a competing supervised algorithm based on a Hidden Markov Model, our unsupervised method demonstrates similar results in the STN detection task and superior results in the DLOR detection task.

Modes of Homogeneous Gradient Flows

no code implementations3 Jul 2020 Ido Cohen, Omri Azencot, Pavel Lifshitz, Guy Gilboa

Definitions for spectrum and filtering are given, and a Parseval-type identity is shown.

Dynamical Systems Computational Engineering, Finance, and Science

Supervised and Unsupervised End-to-End Deep Learning for Gene Ontology Classification of Neural In Situ Hybridization Images

no code implementations30 Nov 2019 Ido Cohen, Eli David, Nathan S. Netanyahu

In recent years, large datasets of high-resolution mammalian neural images have become available, which has prompted active research on the analysis of gene expression data.

Denoising General Classification

Cannot find the paper you are looking for? You can Submit a new open access paper.