AI Paper Atlas

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

Visualizing the Emergence of Intermediate Visual Patterns in DNNs

Mingjie Li, Shaobo Wang, Quanshi Zhang

NeurIPS 2021 · Adversarial robustness and security › Attacks on text and graphs · approach: Biologically inspired robust vision · 0 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
project sample features with the proposed vMF-based visualization instead of PCA, t-SNE, LLE, ISOMAP or DRPRsample-feature visualization · Tiny ImageNet / COCO / CUBPearson correlation between feature strength and classification uncertaintyImproved
compare student DNN distilled at conv_53 against the teacher in number of reliable knowledge pointsVGG-16 student via knowledge distillation · Tiny ImageNetnumber of reliable knowledge pointsHurt
distill from penultimate fully-connected layer instead of conv_53 layer (ablation)VGG-16 student via knowledge distillation · Tiny ImageNetnumber of reliable knowledge points in conv layersHurt
distill from fully-connected layer rather than conv_53, measured at conv_53 (ablation)VGG-16 student via knowledge distillation · Tiny ImageNetfeature similarity to teacher (orientation and strength)Hurt
go from low layer (conv_33) to high layer (conv_53)VGG-16 · Tiny ImageNetratio of reliable knowledge pointsImproved
apply untargeted PGD attack (epsilon 1/255, 20 steps) and compare layersVGG-16 · Tiny ImageNetchange in regional feature orientation and strengthMixed
add alignment loss L_align (alpha=0.1) to similarity loss for regional featuresregional feature projection · Tiny ImageNetqualitative alignment with sample coordinate systemNo clear effect
estimated regional importance w versus Shapley valueimportance estimation · VGG-16 conv_53, multiple datasetssimilarity between w and Shapley valueImproved
relate projected feature strength to estimated importanceregional features h · Tiny ImageNet, COCO, CUBPearson correlation of strength and importanceImproved
choose threshold tau=0.4 for knowledge points (larger vs smaller tau) (ablation)knowledge point counting · VGG-16number of knowledge pointsMixed

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