mlsanityData to 14 Oct 2026

How to Learn when Data Reacts to Your Model: Performative Gradient Descent

arXiv 2102.07698 ↗First posted 15 Feb 2021cs.LG stat.MLcited by 71

Not on the map. Why?

Abstract (arXiv)

Performative distribution shift captures the setting where the choice of which ML model is deployed changes the data distribution. For example, a bank which uses the number of open credit lines to determine a customer's risk of default on a loan may induce customers to open more credit lines in order to improve their chances of being approved. Because of the interactions between the model and data distribution, finding the optimal model parameters is challenging. Works in this area have focused on finding stable points, which can be far from optimal. Here we introduce performative gradient descent (PerfGD), which is the first algorithm which provably converges to the performatively optimal point. PerfGD explicitly captures how changes in the model affects the data distribution and is simple to use. We support our findings with theory and experiments.

Key facts. Cited by 71 papers in the map (data through 14 Oct 2026). Builds on 16 of its 30 references that are in the map. First posted on arXiv 15 Feb 2021.

Builds on

16 of 30 references.

References · 16

  1. 2020Performative Predictioncited by 327
  2. 2015Strategic Classificationcited by 257
  3. 2013Optimal rates for zero-order convex optimization: the power of two function evaluationscited by 198
  4. 2013Non-stationary Stochastic Optimizationcited by 175
  5. 2020Stochastic Optimization for Performative Predictioncited by 93
  6. 2019Strategic Classification is Causal Modeling in Disguisecited by 92
  7. 2020Performative Prediction in a Stateful Worldcited by 90
  8. 2018How Do Classifiers Induce Agents To Invest Effort Strategically?cited by 83
  9. 2020On the Sentence Embeddings from Pre-trained Language Modelscited by 83
  10. 2020Stochastic optimization with decision-dependent distributionscited by 65
  11. 2020Improved Regret for Zeroth-Order Adversarial Bandit Convex Optimisationcited by 33
  12. 2014Optimum Statistical Estimation with Strategic Data Sourcescited by 28
4 more
  1. 2018Strategyproof Linear Regression in High Dimensionscited by 22
  2. 2019Optimal Decision Making Under Strategic Behaviorcited by 20
  3. 2020Manipulation-Proof Machine Learningcited by 20
  4. 2020Gaming Helps! Learning from Strategic Interactions in Natural Dynamicscited by 15

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 5; 2022: 11; 2023: 9; 2024: 20; 2025: 12; 2026: 14.

71 papers.

Sort:

By area:

  1. 2022MetaShift: A Dataset of Datasets for Evaluating Contextual Distribution Shifts and Training Conflictscited by 81
  2. 2022Data Feedback Loops: Model-driven Amplification of Dataset Biasescited by 60
  3. 2021Strategic Classification Made Practicalcited by 59
  4. 2021Who Leads and Who Follows in Strategic Classification?cited by 53
  5. 2023Performative Prediction: Past and Futurecited by 53
  6. 2022Multiplayer Performative Prediction: Learning in Decision-Dependent Gamescited by 44
  7. 2022Performative Powercited by 43
  8. 2022Regret Minimization with Performative Feedbackcited by 40
  9. 2022Anticipating Performativity by Predicting from Predictionscited by 37
  10. 2023Performative Prediction with Neural Networkscited by 34
  11. 2022Multi-agent Performative Prediction: From Global Stability and Optimality to Chaoscited by 33
  12. 2022Making Decisions under Outcome Performativitycited by 33
38 more
  1. 2021How to Learn when Data Gradually Reacts to Your Modelcited by 30
  2. 2023Plug-in Performative Optimizationcited by 22
  3. 2023Performative Federated Learning: A Solution to Model-Dependent and Heterogeneous Distribution Shiftscited by 14
  4. 2024Performative Reinforcement Learning in Gradually Shifting Environmentscited by 14
  5. 2022Stochastic Approximation with Decision-Dependent Distributions: Asymptotic Normality and Optimalitycited by 12
  6. 2023Performative Recommendation: Diversifying Content via Strategic Incentivescited by 12
  7. 2024Non-linear Welfare-Aware Strategic Learningcited by 12
  8. 2021Approximate Regions of Attraction in Learning with Decision-Dependent Distributionscited by 11
  9. 2023Performative Prediction with Bandit Feedback: Learning through Reparameterizationcited by 11
  10. 2024Addressing Polarization and Unfairness in Performative Predictioncited by 11
  11. 2023Zero-Regret Performative Prediction Under Inequality Constraintscited by 10
  12. 2021Model Transferability With Responsive Decision Subjectscited by 9
  13. 2023Incentivizing honest performative predictions with proper scoring rulescited by 9
  14. 2024Automating Data Annotation under Strategic Human Agents: Risks and Potential Solutionscited by 9
  15. 2024Performative Reinforcement Learning with Linear Markov Decision Processcited by 9
  16. 2024Learning under Imitative Strategic Behavior with Unforeseeable Outcomescited by 8
  17. 2025Performative Validity of Recourse Explanationscited by 8
  18. 2024Algorithmic Decision-Making under Agents with Persistent Improvementcited by 7
  19. 2024Performative Prediction on Games and Mechanism Designcited by 7
  20. 2024Distributionally Robust Performative Predictioncited by 7
  21. 2024Learning In Reverse Causal Strategic Environments With Ramifications on Two Sided Marketscited by 6
  22. 2024Algorithmic Fairness in Performative Policy Learning: Escaping the Impossibility of Group Fairnesscited by 6
  23. 2025Decision-Dependent Stochastic Optimization: The Role of Distribution Dynamicscited by 6
  24. 2025On Corruption-Robustness in Performative Reinforcement Learningcited by 6
  25. 2025Performative Policy Gradient: Optimality in Performative Reinforcement Learningcited by 6
  26. 2022Emergent specialization from participation dynamics and multi-learner retrainingcited by 5
  27. 2022Optimizing the Performative Risk under Weak Convexity Assumptionscited by 5
  28. 2023Strategic Distribution Shift of Interacting Agents via Coupled Gradient Flowscited by 5
  29. 2024Clipped SGD Algorithms for Performative Prediction: Tight Bounds for Clipping Bias and Remediescited by 5
  30. 2024Understanding the Training and Generalization of Pretrained Transformer for Sequential Decision Makingcited by 5
  31. 2024Learning the Distribution Map in Reverse Causal Performative Predictioncited by 5
  32. 2025How Strategic Agents Respond: Comparing Analytical Models with LLM-Generated Responses in Strategic Classificationcited by 5
  33. 2024Tight Lower Bounds and Improved Convergence in Performative Predictioncited by 4
  34. 2026The Stability of Online Algorithms in Performative Predictioncited by 4
  35. 2024Practical Performative Policy Learning with Strategic Agentscited by 3
  36. 2024Parameter-Free Algorithms for Performative Regret Minimization under Decision-Dependent Distributionscited by 2
  37. 2024ProFL: Performative Robust Optimal Federated Learningcited by 2
  38. 2024Strategic Conformal Predictioncited by 2

50 most cited of 71 All 71 (JSON)