Noga Bar — accepted AI-conference papers

2025

Revisiting Glorot Initialization for Long-Range Linear Recurrences

Noga Bar, Mariia Seleznova, Yotam Alexander, Gitta Kutyniok, Raja Giryes

NeurIPS 2025poster

Proper initialization is critical for Recurrent Neural Networks (RNNs), particularly in long-range reasoning tasks, where repeated application of the same weight matrix can cause vanishing or exploding signals. A common baseline for linear recurrences is Glorot initialization, designed to ensure sta…

Cited by 0 Source ↗Scholar ↗

2025

ZOQO: Zero-Order Quantized Optimization

Noga Bar, Raja Giryes

ICASSP 2025accepted

The increasing computational and memory demands in deep learning present significant challenges, especially in resource-constrained environments. We introduce a zero-order quantized optimization (ZOQO) method designed for training models with quantized parameters and operations. Our approach leverag…

Cited by 0 Source ↗Scholar ↗