Pith. sign in

REVIEW 6 cited by

How Does Mixup Help With Robustness and Generalization?

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2010.04819 v4 pith:VMKXUTU2 submitted 2020-10-09 cs.LG stat.ML

Linjun Zhang , Zhun Deng , Kenji Kawaguchi , Amirata Ghorbani , James Zou This is my paper
classification cs.LGstat.ML
keywords mixuprobustnessgeneralizationadversarialanalysisaugmentationcorrespondsloss
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Mixup is a popular data augmentation technique based on taking convex combinations of pairs of examples and their labels. This simple technique has been shown to substantially improve both the robustness and the generalization of the trained model. However, it is not well-understood why such improvement occurs. In this paper, we provide theoretical analysis to demonstrate how using Mixup in training helps model robustness and generalization. For robustness, we show that minimizing the Mixup loss corresponds to approximately minimizing an upper bound of the adversarial loss. This explains why models obtained by Mixup training exhibits robustness to several kinds of adversarial attacks such as Fast Gradient Sign Method (FGSM). For generalization, we prove that Mixup augmentation corresponds to a specific type of data-adaptive regularization which reduces overfitting. Our analysis provides new insights and a framework to understand Mixup.

Discussion (0). Sign in to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Bringing Balance to Hand Shape Classification: Mitigating Data Imbalance Through Generative Models

    cs.CV 2025-07 conditional novelty 6.0 of 10

    Pre-training an EfficientNet classifier on GAN-generated balanced hand images, then fine-tuning on real data, raises accuracy on the imbalanced RWTH handshape benchmark from 80.6% to 85.3%.

  2. LangDAug: Langevin Data Augmentation for Multi-Source Domain Generalization in Medical Image Segmentation

    cs.CV 2025-05 conditional novelty 6.0 of 10

    LangDAug generates intermediate-domain images via Langevin dynamics over energy-based models and reports consistent domain generalization gains on two medical segmentation tasks.

  3. Towards Understanding Why Data Augmentation Improves Generalization

    cs.CV 2025-02 conditional novelty 6.0 of 10

    Data augmentation improves generalization through two mechanisms, partial semantic feature removal and feature mixing, which respectively promote diverse and robust feature learning.

  4. ADAPT: A Pseudo-labeling Approach to Combat Concept Drift in Malware Detection

    cs.LG 2025-07 conditional novelty 5.0 of 10

    A pseudo-labeling method with class-specific adaptive thresholds, label-consistent augmentation, and mixup reduces concept-drift performance loss in malware classifiers across five datasets.

  5. RoGA: Towards Generalizable Deepfake Detection through Robust Gradient Alignment

    cs.CV 2025-05 reject novelty 4.0 of 10

    RoGA proposes a SAM-style optimization with per-domain gradient alignment for deepfake detection, but the paper's empirical support is undermined by misreported numbers and target-domain hyperparameter tuning.

  6. Why Pool When You Can Flow? Active Learning with GFlowNets

    cs.LG 2025-08 conditional novelty 3.0 of 10

    Training a GFlowNet to generate high-BALD molecules gives pool-size-independent acquisition and near-BALD classification quality on JAK2 virtual screening.

Pith tools