Alert button
Picture for Levent Sagun

Levent Sagun

Alert button

Networked Inequality: Preferential Attachment Bias in Graph Neural Network Link Prediction

Sep 29, 2023
Arjun Subramonian, Levent Sagun, Yizhou Sun

Viaarxiv icon

Weisfeiler and Lehman Go Measurement Modeling: Probing the Validity of the WL Test

Jul 11, 2023
Arjun Subramonian, Adina Williams, Maximilian Nickel, Yizhou Sun, Levent Sagun

Figure 1 for Weisfeiler and Lehman Go Measurement Modeling: Probing the Validity of the WL Test
Figure 2 for Weisfeiler and Lehman Go Measurement Modeling: Probing the Validity of the WL Test
Figure 3 for Weisfeiler and Lehman Go Measurement Modeling: Probing the Validity of the WL Test
Figure 4 for Weisfeiler and Lehman Go Measurement Modeling: Probing the Validity of the WL Test
Viaarxiv icon

Simplicity Bias Leads to Amplified Performance Disparities

Dec 13, 2022
Samuel J. Bell, Levent Sagun

Figure 1 for Simplicity Bias Leads to Amplified Performance Disparities
Figure 2 for Simplicity Bias Leads to Amplified Performance Disparities
Figure 3 for Simplicity Bias Leads to Amplified Performance Disparities
Figure 4 for Simplicity Bias Leads to Amplified Performance Disparities
Viaarxiv icon

Measuring and signing fairness as performance under multiple stakeholder distributions

Jul 20, 2022
David Lopez-Paz, Diane Bouchacourt, Levent Sagun, Nicolas Usunier

Viaarxiv icon

Understanding out-of-distribution accuracies through quantifying difficulty of test samples

Mar 28, 2022
Berfin Simsek, Melissa Hall, Levent Sagun

Figure 1 for Understanding out-of-distribution accuracies through quantifying difficulty of test samples
Figure 2 for Understanding out-of-distribution accuracies through quantifying difficulty of test samples
Figure 3 for Understanding out-of-distribution accuracies through quantifying difficulty of test samples
Figure 4 for Understanding out-of-distribution accuracies through quantifying difficulty of test samples
Viaarxiv icon

Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision

Feb 22, 2022
Priya Goyal, Quentin Duval, Isaac Seessel, Mathilde Caron, Ishan Misra, Levent Sagun, Armand Joulin, Piotr Bojanowski

Figure 1 for Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision
Figure 2 for Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision
Figure 3 for Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision
Figure 4 for Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision
Viaarxiv icon

Fairness Indicators for Systematic Assessments of Visual Feature Extractors

Feb 15, 2022
Priya Goyal, Adriana Romero Soriano, Caner Hazirbas, Levent Sagun, Nicolas Usunier

Figure 1 for Fairness Indicators for Systematic Assessments of Visual Feature Extractors
Figure 2 for Fairness Indicators for Systematic Assessments of Visual Feature Extractors
Figure 3 for Fairness Indicators for Systematic Assessments of Visual Feature Extractors
Figure 4 for Fairness Indicators for Systematic Assessments of Visual Feature Extractors
Viaarxiv icon

Transformed CNNs: recasting pre-trained convolutional layers with self-attention

Jun 10, 2021
Stéphane d'Ascoli, Levent Sagun, Giulio Biroli, Ari Morcos

Figure 1 for Transformed CNNs: recasting pre-trained convolutional layers with self-attention
Figure 2 for Transformed CNNs: recasting pre-trained convolutional layers with self-attention
Figure 3 for Transformed CNNs: recasting pre-trained convolutional layers with self-attention
Figure 4 for Transformed CNNs: recasting pre-trained convolutional layers with self-attention
Viaarxiv icon

ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases

Mar 19, 2021
Stéphane d'Ascoli, Hugo Touvron, Matthew Leavitt, Ari Morcos, Giulio Biroli, Levent Sagun

Figure 1 for ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases
Figure 2 for ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases
Figure 3 for ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases
Figure 4 for ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases
Viaarxiv icon