Crowding Out the Noise: Algorithmic Collective Action under Differential Privacy

Keyphrases

• Algorithmic Harms 20%
• Average Participation 20%
• Average Utility 20%
• Differentially Private Stochastic Gradient Descent 20%
• Grassroots 20%
• Individual Protection 20%
• Participation Costs 20%
• Privacy Parameters 20%
• Private Model 20%
• Private Training 20%
• Regulatory Focus 20%
• Social Data 20%
• Social Inequity 20%
• Training Regime 20%
• Trustworthy AI Systems 20%
• Utility Costs 20%

Social Sciences

• Cell Excitability 14%
• Crowding Out 100%
• Data Protection 14%
• Differential Privacy 100%
• Social Data 14%

Computer Science

• Average Utility 16%
• Trustworthiness Model 16%