Picture for Levent Sagun

Levent Sagun

A Differentiable Rank-Based Objective For Better Feature Learning

Add code
Feb 13, 2025
Viaarxiv icon

Chained Tuning Leads to Biased Forgetting

Add code
Dec 21, 2024
Figure 1 for Chained Tuning Leads to Biased Forgetting
Figure 2 for Chained Tuning Leads to Biased Forgetting
Figure 3 for Chained Tuning Leads to Biased Forgetting
Figure 4 for Chained Tuning Leads to Biased Forgetting
Viaarxiv icon

On the Role of Speech Data in Reducing Toxicity Detection Bias

Add code
Nov 12, 2024
Figure 1 for On the Role of Speech Data in Reducing Toxicity Detection Bias
Figure 2 for On the Role of Speech Data in Reducing Toxicity Detection Bias
Figure 3 for On the Role of Speech Data in Reducing Toxicity Detection Bias
Figure 4 for On the Role of Speech Data in Reducing Toxicity Detection Bias
Viaarxiv icon

The Root Shapes the Fruit: On the Persistence of Gender-Exclusive Harms in Aligned Language Models

Add code
Nov 06, 2024
Figure 1 for The Root Shapes the Fruit: On the Persistence of Gender-Exclusive Harms in Aligned Language Models
Figure 2 for The Root Shapes the Fruit: On the Persistence of Gender-Exclusive Harms in Aligned Language Models
Figure 3 for The Root Shapes the Fruit: On the Persistence of Gender-Exclusive Harms in Aligned Language Models
Figure 4 for The Root Shapes the Fruit: On the Persistence of Gender-Exclusive Harms in Aligned Language Models
Viaarxiv icon

Reassessing the Validity of Spurious Correlations Benchmarks

Add code
Sep 06, 2024
Figure 1 for Reassessing the Validity of Spurious Correlations Benchmarks
Figure 2 for Reassessing the Validity of Spurious Correlations Benchmarks
Figure 3 for Reassessing the Validity of Spurious Correlations Benchmarks
Figure 4 for Reassessing the Validity of Spurious Correlations Benchmarks
Viaarxiv icon

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

Add code
Sep 29, 2023
Figure 1 for Networked Inequality: Preferential Attachment Bias in Graph Neural Network Link Prediction
Figure 2 for Networked Inequality: Preferential Attachment Bias in Graph Neural Network Link Prediction
Figure 3 for Networked Inequality: Preferential Attachment Bias in Graph Neural Network Link Prediction
Figure 4 for Networked Inequality: Preferential Attachment Bias in Graph Neural Network Link Prediction
Viaarxiv icon

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

Add code
Jul 11, 2023
Viaarxiv icon

Simplicity Bias Leads to Amplified Performance Disparities

Add code
Dec 13, 2022
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

Add code
Jul 20, 2022
Viaarxiv icon

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

Add code
Mar 28, 2022
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