Picture for Piotr Zielinski

Piotr Zielinski

A Closer Look at Hardware-Friendly Weight Quantization

Add code
Oct 07, 2022
Figure 1 for A Closer Look at Hardware-Friendly Weight Quantization
Figure 2 for A Closer Look at Hardware-Friendly Weight Quantization
Figure 3 for A Closer Look at Hardware-Friendly Weight Quantization
Figure 4 for A Closer Look at Hardware-Friendly Weight Quantization
Viaarxiv icon

On the Generalization Mystery in Deep Learning

Add code
Mar 31, 2022
Figure 1 for On the Generalization Mystery in Deep Learning
Figure 2 for On the Generalization Mystery in Deep Learning
Figure 3 for On the Generalization Mystery in Deep Learning
Figure 4 for On the Generalization Mystery in Deep Learning
Viaarxiv icon

Enabling Binary Neural Network Training on the Edge

Add code
Feb 10, 2021
Figure 1 for Enabling Binary Neural Network Training on the Edge
Figure 2 for Enabling Binary Neural Network Training on the Edge
Figure 3 for Enabling Binary Neural Network Training on the Edge
Figure 4 for Enabling Binary Neural Network Training on the Edge
Viaarxiv icon

Making Coherence Out of Nothing At All: Measuring the Evolution of Gradient Alignment

Add code
Aug 03, 2020
Figure 1 for Making Coherence Out of Nothing At All: Measuring the Evolution of Gradient Alignment
Figure 2 for Making Coherence Out of Nothing At All: Measuring the Evolution of Gradient Alignment
Figure 3 for Making Coherence Out of Nothing At All: Measuring the Evolution of Gradient Alignment
Figure 4 for Making Coherence Out of Nothing At All: Measuring the Evolution of Gradient Alignment
Viaarxiv icon

Explaining Memorization and Generalization: A Large-Scale Study with Coherent Gradients

Add code
Mar 16, 2020
Figure 1 for Explaining Memorization and Generalization: A Large-Scale Study with Coherent Gradients
Figure 2 for Explaining Memorization and Generalization: A Large-Scale Study with Coherent Gradients
Figure 3 for Explaining Memorization and Generalization: A Large-Scale Study with Coherent Gradients
Figure 4 for Explaining Memorization and Generalization: A Large-Scale Study with Coherent Gradients
Viaarxiv icon