Performative Power

Institute Homepage

Institute Homepage Sign In

Back

Research Overview

Performativity in machine learning

Measuring performative power of online search

Performative Prediction: Past and Future

Applications of algorithmic collective action

Living artefact of collective action cases

Large language models and surveys

Questioning the survey responses of LLMs

Folktexts

Evaluating LLMs as risk scores

Social Foundations of Computation Members Publications

Performative Power

Performative power is a new notion of power that quantifies how much a firm can steer a target population through its algorithmic actions. We show that the definition enjoys many desirable theoretical properties, while proving identifiability from observational causal inference [].

Members

Social Foundations of Computation

Moritz Hardt

Director

Algorithms and Society

Celestine Mendler-Dünner

Hector Endowed Fellow of the ELLIS Institute

Publications

Social Foundations of Computation Algorithms and Society Conference Paper Performative Power Hardt, M., Jagadeesan, M., Mendler-Dünner, C. In Advances in Neural Information Processing Systems 35 (NeurIPS 2022), Curran Associates Inc., The Thirty-Six Annual Conference on Neural Information Processing Systems (NeurIPS), March 2022 (Published) URL BibTeX