mlsanityData to 7 Oct 2026

Pruning has a disparate impact on model accuracy

arXiv 2205.13574 ↗First posted 26 May 2022cs.LGcited by 43

Not on the map. Why?

Abstract (arXiv)

Network pruning is a widely-used compression technique that is able to significantly scale down overparameterized models with minimal loss of accuracy. This paper shows that pruning may create or exacerbate disparate impacts. The paper sheds light on the factors to cause such disparities, suggesting differences in gradient norms and distance to decision boundary across groups to be responsible for this critical issue. It analyzes these factors in detail, providing both theoretical and empirical support, and proposes a simple, yet effective, solution that mitigates the disparate impacts caused by pruning.

Key facts. Cited by 43 papers in the map (data through 7 Oct 2026). Builds on 35 of its 56 references that are in the map. First posted on arXiv 26 May 2022.

Builds on

35 of 56 references.

References · 35

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  3. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  4. 2015Learning both Weights and Connections for Efficient Neural Networkscited by 1,998
  5. 2019A Survey on Bias and Fairness in Machine Learningcited by 1,911
  6. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,324
  7. 2020What is the State of Neural Network Pruning?cited by 578
  8. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 439
  9. 2017Age Progression/Regression by Conditional Adversarial Autoencodercited by 426
  10. 2020Fairness in Machine Learning: A Surveycited by 369
  11. 2019Differential Privacy Has Disparate Impact on Model Accuracycited by 316
  12. 2020Comparing Rewinding and Fine-tuning in Neural Network Pruningcited by 238
23 more
  1. 2020Fairness in Machine Learningcited by 194
  2. 2019What Do Compressed Deep Neural Networks Forget?cited by 164
  3. 2020Characterising Bias in Compressed Modelscited by 162
  4. 2019Inherent Tradeoffs in Learning Fair Representationscited by 144
  5. 2020To be Robust or to be Fair: Towards Fairness in Adversarial Trainingcited by 137
  6. 2018A Systematic DNN Weight Pruning Framework using Alternating Direction Method of Multiplierscited by 118
  7. 2020Lagrangian Duality for Constrained Deep Learningcited by 72
  8. 2018Sparse DNNs with Improved Adversarial Robustnesscited by 71
  9. 2019Fair Decision Making using Privacy-Protected Datacited by 58
  10. 2022Differential Privacy and Fairness in Decisions and Learning Tasks: A Surveycited by 47
  11. 2021The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translationcited by 46
  12. 2021Beyond Preserved Accuracy: Evaluating Loyalty and Robustness of BERT Compressioncited by 38
  13. 2021Decision Making with Differential Privacy under a Fairness Lenscited by 32
  14. 2021DP-SGD vs PATE: Which Has Less Disparate Impact on Model Accuracy?cited by 32
  15. 2021The Rich Get Richer: Disparate Impact of Semi-Supervised Learningcited by 28
  16. 2019Towards Compact and Robust Deep Neural Networkscited by 23
  17. 2020Prune Responsiblycited by 20
  18. 2022Can Model Compression Improve NLP Fairnesscited by 18
  19. 2019SiPPing Neural Networks: Sensitivity-informed Provable Pruning of Neural Networkscited by 17
  20. 2020Going Beyond Classification Accuracy Metrics in Model Compressioncited by 15
  21. 2020Bias and Variance of Post-processing in Differential Privacycited by 14
  22. 2021Simon Says: Evaluating and Mitigating Bias in Pruned Neural Networks with Knowledge Distillationcited by 14
  23. 2022Tiny, always-on and fragile: Bias propagation through design choices in on-device machine learning workflowscited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 2; 2023: 9; 2024: 11; 2025: 10; 2026: 11.

43 papers.

Sort:

By area:

  1. 2023Sparse Model Soups: A Recipe for Improved Pruning via Model Averagingcited by 20
  2. 2024Superficial Safety Alignment Hypothesiscited by 14
  3. 2022Recall Distortion in Neural Network Pruning and the Undecayed Pruning Algorithmcited by 9
  4. 2023Getting Away with More Network Pruning: From Sparsity to Geometry and Linear Regionscited by 8
  5. 2024On Fairness of Low-Rank Adaptation of Large Modelscited by 8
  6. 2023Quantifying and mitigating the impact of label errors on model disparity metricscited by 7
  7. 2023Balancing Act: Constraining Disparate Impact in Sparse Modelscited by 7
  8. 2024DRoP: Distributionally Robust Data Pruningcited by 7
  9. 2024LayerMerge: Neural Network Depth Compression through Layer Pruning and Mergingcited by 7
  10. 2025Zero-shot Quantization: A Comprehensive Surveycited by 7
  11. 2024BMFT: Achieving Fairness via Bias-based Weight Masking Fine-tuningcited by 6
  12. 2022Fairness Increases Adversarial Vulnerabilitycited by 5
31 more
  1. 2024Position: Cracking the Code of Cascading Disparity Towards Marginalized Communitiescited by 4
  2. 2023FairDP: Certified Fairness with Differential Privacycited by 3
  3. 2023Uncovering the Hidden Cost of Model Compressioncited by 3
  4. 2025Fair Foundation Models for Medical Image Analysis: Challenges and Perspectivescited by 3
  5. 2023On The Fairness Impacts of Hardware Selection in Machine Learningcited by 2
  6. 2024FairLoRA: Unpacking Bias Mitigation in Vision Models with Fairness-Driven Low-Rank Adaptationcited by 2
  7. 2024You Never Know: Quantization Induces Inconsistent Biases in Vision-Language Foundation Modelscited by 2
  8. 2025An Efficient Training Algorithm for Models with Block-wise Sparsitycited by 2
  9. 2025Explaining How Quantization Disparately Skews a Modelcited by 2
  10. 2025Building Trustworthy AI for Materials Discovery: From Autonomous Laboratories to Z-scorescited by 2
  11. 2026Weight Pruning Amplifies Bias: A Multi-Method Study of Compressed LLMs for Edge AIcited by 2
  12. 2023Samples on Thin Ice: Re-Evaluating Adversarial Pruning of Neural Networkscited by 1
  13. 2024CurvFed: Curvature-Aligned Federated Learning for Fairness without Demographicscited by 1
  14. 2024Fisher-aware Quantization for DETR Detectors with Critical-category Objectivescited by 1
  15. 2025MicroNAS: An Automated Framework for Developing a Fall Detection Systemcited by 1
  16. 2026Decoupling Generalizability and Membership Privacy Risks in Neural Networkscited by 1
  17. 2026SynQ: Accurate Zero-shot Quantization by Synthesis-aware Fine-tuningcited by 1
  18. 2023Integrating Fairness and Model Pruning Through Bi-level Optimizationcited by 0
  19. 2024Differentially Private Data Release on Graphs: Inefficiencies and Unfairnesscited by 0
  20. 2025As easy as PIE: understanding when pruning causes language models to disagreecited by 0
  21. 2025NuWa: Deriving Lightweight Class-Specific Vision Transformers for Edge Devicescited by 0
  22. 2025MGS: Markov Greedy Sums for Accurate Low-Bitwidth Floating-Point Accumulationcited by 0
  23. 2025Optimization over Trained (and Sparse) Neural Networks: A Surrogate within a Surrogatecited by 0
  24. 2026TraceNAS: Zero-shot LLM Pruning via Gradient Trace Correlationcited by 0
  25. 2026Learnability and Privacy Vulnerability are Entangled in a Few Critical Weightscited by 0
  26. 2026Ethical Fairness in Ubiquitous Health Sensing without Known Attributescited by 0
  27. 2026Fair Finetuning Mitigates Distribution Inference Attackscited by 0
  28. 2026Lottery Tickets Are Not Deployment Ticketscited by 0
  29. 2026The Asymmetric Harms of LLM Compressioncited by 0
  30. 2026Input-Layer Starvation: Why Per-Layer Pruning Breaks IoT Intrusion Detectorscited by 0
  31. 2026Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMscited by 0