mlsanityData to 7 Oct 2026

Performative Prediction

arXiv 2002.06673 ↗First posted 16 Feb 2020cs.LG cs.GT stat.MLcited by 318

Not on the map. Why?

Abstract (arXiv)

When predictions support decisions they may influence the outcome they aim to predict. We call such predictions performative; the prediction influences the target. Performativity is a well-studied phenomenon in policy-making that has so far been neglected in supervised learning. When ignored, performativity surfaces as undesirable distribution shift, routinely addressed with retraining. We develop a risk minimization framework for performative prediction bringing together concepts from statistics, game theory, and causality. A conceptual novelty is an equilibrium notion we call performative stability. Performative stability implies that the predictions are calibrated not against past outcomes, but against the future outcomes that manifest from acting on the prediction. Our main results are necessary and sufficient conditions for the convergence of retraining to a performatively stable point of nearly minimal loss. In full generality, performative prediction strictly subsumes the setting known as strategic classification. We thus also give the first sufficient conditions for retraining to overcome strategic feedback effects.

Key facts. Cited by 318 papers in the map (data through 7 Oct 2026). First posted on arXiv 16 Feb 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 23; 2022: 44; 2023: 50; 2024: 59; 2025: 56; 2026: 86.

318 papers.

Sort:

By area:

  1. 2022Online Decision Transformercited by 194
  2. 2021The Values Encoded in Machine Learning Researchcited by 172
  3. 2023On the Stability of Iterative Retraining of Generative Models on their own Datacited by 98
  4. 2021Outside the Echo Chamber: Optimizing the Performative Riskcited by 95
  5. 2021How to Learn when Data Reacts to Your Model: Performative Gradient Descentcited by 70
  6. 2024Feedback Loops With Language Models Drive In-Context Reward Hackingcited by 70
  7. 2021Strategic Classification in the Darkcited by 66
  8. 2022Data Feedback Loops: Model-driven Amplification of Dataset Biasescited by 59
  9. 2021Strategic Classification Made Practicalcited by 58
  10. 2024Fairness Feedback Loops: Training on Synthetic Data Amplifies Biascited by 55
  11. 2021Who Leads and Who Follows in Strategic Classification?cited by 53
  12. 2022Conformal Prediction Under Feedback Covariate Shift for Biomolecular Designcited by 53
38 more
  1. 2022Leveraging Explanations in Interactive Machine Learning: An Overviewcited by 52
  2. 2023Performative Prediction: Past and Futurecited by 52
  3. 2021Information Discrepancy in Strategic Learningcited by 48
  4. 2021Alternative Microfoundations for Strategic Classificationcited by 47
  5. 2023Evaluating Superhuman Models with Consistency Checkscited by 47
  6. 2022Multiplayer Performative Prediction: Learning in Decision-Dependent Gamescited by 42
  7. 2022Performative Powercited by 42
  8. 2021Sample-Efficient Safety Assurances using Conformal Predictioncited by 41
  9. 2022Regret Minimization with Performative Feedbackcited by 40
  10. 2022Turning the Tables: Biased, Imbalanced, Dynamic Tabular Datasets for ML Evaluationcited by 39
  11. 2022Anticipating Performativity by Predicting from Predictionscited by 37
  12. 2022Modeling Content Creator Incentives on Algorithm-Curated Platformscited by 35
  13. 2022The Sample Complexity of Online Contract Designcited by 35
  14. 2023Difficult Lessons on Social Prediction from Wisconsin Public Schoolscited by 35
  15. 2024Self-Correcting Self-Consuming Loops for Generative Model Trainingcited by 34
  16. 2022Multi-agent Performative Prediction: From Global Stability and Optimality to Chaoscited by 33
  17. 2023Performative Prediction with Neural Networkscited by 33
  18. 2024AI Alignment with Changing and Influenceable Reward Functionscited by 33
  19. 2021Stateful Strategic Regressioncited by 32
  20. 2021Strategic Instrumental Variable Regression: Recovering Causal Relationships From Strategic Responsescited by 32
  21. 2022Making Decisions under Outcome Performativitycited by 32
  22. 2024Assessing Political Bias in Large Language Modelscited by 32
  23. 2022Decision-Dependent Risk Minimization in Geometrically Decaying Dynamic Environmentscited by 31
  24. 2022A Validity Perspective on Evaluating the Justified Use of Data-driven Decision-making Algorithmscited by 30
  25. 2023Causal Strategic Classification: A Tale of Two Shiftscited by 30
  26. 2021How to Learn when Data Gradually Reacts to Your Modelcited by 29
  27. 2022Preference Dynamics Under Personalized Recommendationscited by 25
  28. 2022Performative Reinforcement Learningcited by 25
  29. 2023Long-Term Fairness with Unknown Dynamicscited by 25
  30. 2022Learning Losses for Strategic Classificationcited by 22
  31. 2022General Intelligence Requires Rethinking Explorationcited by 20
  32. 2023Plug-in Performative Optimizationcited by 20
  33. 2021Zeroth-Order Methods for Convex-Concave Minmax Problems: Applications to Decision-Dependent Risk Minimizationcited by 18
  34. 2022Achieving Long-Term Fairness in Sequential Decision Makingcited by 18
  35. 2022The worst of both worlds: A comparative analysis of errors in learning from data in psychology and machine learningcited by 17
  36. 2023Fundamental Bounds on Online Strategic Classificationcited by 17
  37. 2023Fine-Tuning Games: Bargaining and Adaptation for General-Purpose Modelscited by 17
  38. 2022Human-AI Collaboration in Decision-Making: Beyond Learning to Defercited by 16

50 most cited of 318 All 318 (JSON)