Media Summary: Francesco Laccone (ISTI-CNR, FAIR Spoke 8 - Pervasive AI) presents " Disclaimer: This video is generated with Google's NotebookLM. Spectral Wasserstein From Gradient Descent to Adam. Here are some

Optimizer Geometry In Modern Deep - Detailed Analysis & Overview

Francesco Laccone (ISTI-CNR, FAIR Spoke 8 - Pervasive AI) presents " Disclaimer: This video is generated with Google's NotebookLM. Spectral Wasserstein From Gradient Descent to Adam. Here are some Dive into the fascinating world of Polyhedral This is the video for our paper "GeoCalib: Learning Single-image Calibration with In this video, you'll learn how Adam makes gradient descent faster, smoother, and more reliable by combining the strengths of ...

This video summarizes a new research paper: MARS-M: When Variance Reduction Meets Matrices LLM Training Revolution: ... From physics-informed neural networks that struggle when equations become tightly coupled, to fresh stability theory explaining ...

Photo Gallery

"Optimizer Geometry in Modern Deep Learning" – Zhiyuan Li, Research at TTIC
Optimization for Deep Learning (Momentum, RMSprop, AdaGrad, Adam)
Geometric deep learning for constrained shape optimization of triangulated architectural surfaces
Podcast - A New Geometry of Learning
Who's Adam and What's He Optimizing? | Deep Dive into Optimizers for Machine Learning!
A New Geometry of Learning
Optimizers - EXPLAINED!
Polyhedral Optimization Explained: Geometry of Code
[ECCV 2024] GeoCalib: Learning Single-image Calibration with Geometric Optimization
This Simple Optimizer Is Revolutionizing How We Train AI [Muon]
ADAM Optimization Algorithm Explained Visually | Deep Learning #13
MARS-M Revealed: The New Matrix Optimizer That Speeds Up LLM Training
View Detailed Profile
"Optimizer Geometry in Modern Deep Learning" – Zhiyuan Li, Research at TTIC

"Optimizer Geometry in Modern Deep Learning" – Zhiyuan Li, Research at TTIC

Optimizer Geometry in Modern Deep

Optimization for Deep Learning (Momentum, RMSprop, AdaGrad, Adam)

Optimization for Deep Learning (Momentum, RMSprop, AdaGrad, Adam)

Here we cover six

Geometric deep learning for constrained shape optimization of triangulated architectural surfaces

Geometric deep learning for constrained shape optimization of triangulated architectural surfaces

Francesco Laccone (ISTI-CNR, FAIR Spoke 8 - Pervasive AI) presents "

Podcast - A New Geometry of Learning

Podcast - A New Geometry of Learning

Disclaimer: This video is generated with Google's NotebookLM. https://arxiv.org/pdf/2604.04891 Spectral Wasserstein

Who's Adam and What's He Optimizing? | Deep Dive into Optimizers for Machine Learning!

Who's Adam and What's He Optimizing? | Deep Dive into Optimizers for Machine Learning!

Welcome to our

A New Geometry of Learning

A New Geometry of Learning

Disclaimer: This video is generated with Google's NotebookLM. https://arxiv.org/pdf/2604.04891 Spectral Wasserstein

Optimizers - EXPLAINED!

Optimizers - EXPLAINED!

From Gradient Descent to Adam. Here are some

Polyhedral Optimization Explained: Geometry of Code

Polyhedral Optimization Explained: Geometry of Code

Dive into the fascinating world of Polyhedral

[ECCV 2024] GeoCalib: Learning Single-image Calibration with Geometric Optimization

[ECCV 2024] GeoCalib: Learning Single-image Calibration with Geometric Optimization

This is the video for our paper "GeoCalib: Learning Single-image Calibration with

This Simple Optimizer Is Revolutionizing How We Train AI [Muon]

This Simple Optimizer Is Revolutionizing How We Train AI [Muon]

The Muon

ADAM Optimization Algorithm Explained Visually | Deep Learning #13

ADAM Optimization Algorithm Explained Visually | Deep Learning #13

In this video, you'll learn how Adam makes gradient descent faster, smoother, and more reliable by combining the strengths of ...

MARS-M Revealed: The New Matrix Optimizer That Speeds Up LLM Training

MARS-M Revealed: The New Matrix Optimizer That Speeds Up LLM Training

This video summarizes a new research paper: MARS-M: When Variance Reduction Meets Matrices LLM Training Revolution: ...

Spectral Optimizers and Stability: Robust Learning Theory for Modern Training

Spectral Optimizers and Stability: Robust Learning Theory for Modern Training

From physics-informed neural networks that struggle when equations become tightly coupled, to fresh stability theory explaining ...