# Stress-Testing ML Pipelines with Adversarial Data Corruption

arXiv: https://arxiv.org/abs/2506.01230 · HTML: https://mlsanity.com/paper/2506.01230/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 54 references that are in the map.
- First posted on arXiv 2 Jun 2025.

## Builds on (most cited first, 20 of 20)

- [Equality of Opportunity in Supervised Learning](https://mlsanity.com/paper/1610.02413/) (2016)
- [Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning](https://mlsanity.com/paper/1712.05526/) (2017)
- [Distribution-Free Predictive Inference For Regression](https://mlsanity.com/paper/1604.04173/) (2016)
- [Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networks](https://mlsanity.com/paper/1804.00792/) (2018)
- [Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization](https://mlsanity.com/paper/1708.08689/) (2017)
- [Just How Toxic is Data Poisoning? A Unified Benchmark for Backdoor and Data Poisoning Attacks](https://mlsanity.com/paper/2006.12557/) (2020)
- [Wild Patterns Reloaded: A Survey of Machine Learning Security against Training Data Poisoning](https://mlsanity.com/paper/2205.01992/) (2022)
- [Recent Advances in Bayesian Optimization](https://mlsanity.com/paper/2206.03301/) (2022)
- [Subpopulation Data Poisoning Attacks](https://mlsanity.com/paper/2006.14026/) (2020)
- [Sample Selection for Fair and Robust Training](https://mlsanity.com/paper/2110.14222/) (2021)
- [Interpretable Data-Based Explanations for Fairness Debugging](https://mlsanity.com/paper/2112.09745/) (2021)
- [Indiscriminate Data Poisoning Attacks on Neural Networks](https://mlsanity.com/paper/2204.09092/) (2022)
- [BoostClean: Automated Error Detection and Repair for Machine Learning](https://mlsanity.com/paper/1711.01299/) (2017)
- [Exploring the Limits of Model-Targeted Indiscriminate Data Poisoning Attacks](https://mlsanity.com/paper/2303.03592/) (2023)
- [Improving Fair Training under Correlation Shifts](https://mlsanity.com/paper/2302.02323/) (2023)
- [Through the Data Management Lens: Experimental Analysis and Evaluation of Fair Classification](https://mlsanity.com/paper/2101.07361/) (2021)
- [REIN: A Comprehensive Benchmark Framework for Data Cleaning Methods in ML Pipelines](https://mlsanity.com/paper/2302.04702/) (2023)
- [DiffPrep: Differentiable Data Preprocessing Pipeline Search for Learning over Tabular Data](https://mlsanity.com/paper/2308.10915/) (2023)
- [Explaining Inference Queries with Bayesian Optimization](https://mlsanity.com/paper/2102.05308/) (2021)
- [Investigating Labeler Bias in Face Annotation for Machine Learning](https://mlsanity.com/paper/2301.09902/) (2023)

## Built on by (most cited first, 1 of 1)

- [QuAIL: Quality-Aware Inertial Learning for Robust Training under Data Corruption](https://mlsanity.com/paper/2602.03686/) (2026)

## Abstract (arXiv)

Structured data-quality issues, such as missing values correlated with demographics, culturally biased labels, or systemic selection biases, routinely degrade the reliability of machine-learning pipelines. Regulators now increasingly demand evidence that high-stakes systems can withstand these realistic, interdependent errors, yet current robustness evaluations typically use random or overly simplistic corruptions, leaving worst-case scenarios unexplored. We introduce SAVAGE, a causally inspired framework that (i) formally models realistic data-quality issues through dependency graphs and flexible corruption templates, and (ii) systematically discovers corruption patterns that maximally degrade a target performance metric. SAVAGE employs a bi-level optimization approach to efficiently identify vulnerable data subpopulations and fine-tune corruption severity, treating the full ML pipeline, including preprocessing and potentially non-differentiable models, as a black box. Extensive experiments across multiple datasets and ML tasks (data cleaning, fairness-aware learning, uncertainty quantification) demonstrate that even a small fraction (around 5 %) of structured corruptions identified by SAVAGE severely impacts model performance, far exceeding random or manually crafted errors, and invalidating core assumptions of existing techniques. Thus, SAVAGE provides a practical tool for rigorous pipeline stress-testing, a benchmark for evaluating robustness methods, and actionable guidance for designing more resilient data workflows.