AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Dataset Inference: Ownership Resolution in Machine Learning

Pratyush Maini, Mohammad Yaghini, Nicolas Papernot

ICLR 2021 · Adversarial robustness and security › Attacks on text and graphs · approach: Membership inference and inversion · 23 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR10, data distillationeffect size (Δμ): 0.586 -> 0.778Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR10, different architectureeffect size (Δμ): 0.645 -> 1.400Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR10, zero-shot learningeffect size (Δμ): 0.371 -> 0.406Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR10, fine-tuningeffect size (Δμ): 0.832 -> 1.839Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR10, label-query extractioneffect size (Δμ): 0.475 -> 1.006Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR10, logit-query extractioneffect size (Δμ): 0.563 -> 1.048Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR100, data distillationeffect size (Δμ): 0.362 -> 1.098Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR100, different architectureeffect size (Δμ): 1.016 -> 1.471Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR100, zero-shot learningeffect size (Δμ): 0.466 -> 0.405Mixed
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR100, fine-tuningeffect size (Δμ): 1.047 -> 1.423Improved
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR100, label-query extractioneffect size (Δμ): 0.270 -> 0.107Mixed
Use Blind Walk instead of MinGD for embedding generationDataset inference · CIFAR100, logit-query extractioneffect size (Δμ): 0.385 -> 0.184Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy