# The Limitations of Model Retraining in the Face of Performativity

arXiv: https://arxiv.org/abs/2408.08499 · HTML: https://mlsanity.com/paper/2408.08499/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 18 references that are in the map.
- First posted on arXiv 16 Aug 2024.

## Builds on (most cited first, 8 of 8)

- [Performative Prediction](https://mlsanity.com/paper/2002.06673/) (2020)
- [Outside the Echo Chamber: Optimizing the Performative Risk](https://mlsanity.com/paper/2102.08570/) (2021)
- [Performative Prediction in a Stateful World](https://mlsanity.com/paper/2011.03885/) (2020)
- [Data Feedback Loops: Model-driven Amplification of Dataset Biases](https://mlsanity.com/paper/2209.03942/) (2022)
- [Performative Prediction: Past and Future](https://mlsanity.com/paper/2310.16608/) (2023)
- [Regret Minimization with Performative Feedback](https://mlsanity.com/paper/2202.00628/) (2022)
- [On classification of strategic agents who can both game and improve](https://mlsanity.com/paper/2203.00124/) (2022)
- [The Limits and Potentials of Local SGD for Distributed Heterogeneous Learning with Intermittent Communication](https://mlsanity.com/paper/2405.11667/) (2024)

## Built on by (most cited first, 4 of 4)

- [Dissecting Performative Prediction: A Comprehensive Survey](https://mlsanity.com/paper/2602.10176/) (2026)
- [Strategically Deceptive Model Deployment in Performative Prediction](https://mlsanity.com/paper/2506.09044/) (2025)
- [The Challenger: When Do New Data Sources Justify Switching Machine Learning Models?](https://mlsanity.com/paper/2512.18390/) (2025)
- [When Predictions Shape Reality: A Socio-Technical Synthesis of Performative Predictions in Machine Learning](https://mlsanity.com/paper/2601.04447/) (2026)

## Abstract (arXiv)

We study stochastic optimization in the context of performative shifts, where the data distribution changes in response to the deployed model. We demonstrate that naive retraining can be provably suboptimal even for simple distribution shifts. The issue worsens when models are retrained given a finite number of samples at each retraining step. We show that adding regularization to retraining corrects both of these issues, attaining provably optimal models in the face of distribution shifts. Our work advocates rethinking how machine learning models are retrained in the presence of performative effects.