Corpus ID: 231879702

Defuse: Harnessing Unrestricted Adversarial Examples for Debugging Models Beyond Test Accuracy

@article{Slack2021DefuseHU,
  title={Defuse: Harnessing Unrestricted Adversarial Examples for Debugging Models Beyond Test Accuracy},
  author={Dylan Slack and N. Rauschmayr and K. Kenthapadi},
  journal={ArXiv},
  year={2021},
  volume={abs/2102.06162}
}
We typically compute aggregate statistics on held-out test data to assess the generalization of machine learning models. However, statistics on test data often overstate model generalization, and thus, the performance of deployed machine learning models can be variable and untrustworthy. Motivated by these concerns, we develop methods to automatically discover and correct model errors beyond those available in the data. We propose Defuse, a method that generates novel model misclassifications… Expand

References

SHOWING 1-10 OF 39 REFERENCES
Bayes-TrEx: Model Transparency by Example
  • 2
Generating Natural Adversarial Examples
  • 275
  • Highly Influential
  • PDF
TensorFuzz: Debugging Neural Networks with Coverage-Guided Fuzzing
  • 123
  • PDF
Training Set Debugging Using Trusted Items
  • 33
  • PDF
Flipper: A Systematic Approach to Debugging Training Sets
  • 14
  • PDF
Constructing Unrestricted Adversarial Examples with Generative Models
  • 120
  • Highly Influential
  • PDF
Evaluating Models' Local Decision Boundaries via Contrast Sets
  • 32
  • PDF
Model Assertions for Debugging Machine Learning
  • 16
  • PDF
Socratic Learning: Augmenting Generative Models to Incorporate Latent Subsets in Training Data
  • 21
  • PDF
Data Augmentation Generative Adversarial Networks
  • 404
  • PDF
...
1
2
3
4
...