REVIEW 2 cited by
Understanding the Limitations of Conditional Generative Models
Not yet reviewed by Pith; the record is open.
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
read the original abstract
Class-conditional generative models hold promise to overcome the shortcomings of their discriminative counterparts. They are a natural choice to solve discriminative tasks in a robust manner as they jointly optimize for predictive performance and accurate modeling of the input distribution. In this work, we investigate robust classification with likelihood-based generative models from a theoretical and practical perspective to investigate if they can deliver on their promises. Our analysis focuses on a spectrum of robustness properties: (1) Detection of worst-case outliers in the form of adversarial examples; (2) Detection of average-case outliers in the form of ambiguous inputs and (3) Detection of incorrectly labeled in-distribution inputs. Our theoretical result reveals that it is impossible to guarantee detectability of adversarially-perturbed inputs even for near-optimal generative classifiers. Experimentally, we find that while we are able to train robust models for MNIST, robustness completely breaks down on CIFAR10. We relate this failure to various undesirable model properties that can be traced to the maximum likelihood training objective. Despite being a common choice in the literature, our results indicate that likelihood-based conditional generative models may are surprisingly ineffective for robust classification.
Forward citations
Cited by 2 Pith papers
-
ControlTac: Force- and Position-Controlled Tactile Data Augmentation with a Single Reference Image
A two-stage diffusion framework generates realistic tactile images from one reference image, conditioned on target contact force and position, and the generated images improve downstream force estimation, pose estimat...
-
Pretrained Reversible Generation as Unsupervised Visual Representation Learning
Reversing a pretrained flow or diffusion generator and fine-tuning it with a classification head yields strong image classifiers, reaching 78.1% top-1 on ImageNet-64.
Discussion (0). Continue with ORCID to comment.