AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Labeling Neural Representations with Inverse Recognition

Kirill Bykov, Laura Kopf, Shinichi Nakajima, Marius Kloft, Marina MC Höhne

NeurIPS 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Self-explaining and visual explanation networks · 1 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 INVERT instead of Network Dissection (label-only vs segmentation masks)neuron labeling · ImageNet ResNet18 output neuronsexplanation accuracy: 19.2% -> 73.2%Improved
use INVERT instead of Network Dissectionneuron labeling · ImageNet GoogleNet output neuronsexplanation accuracy: 19.7% -> 82.2%Improved
use INVERT instead of Network Dissectionneuron labeling · ImageNet DenseNet161 output neuronsexplanation accuracy: 19.1% -> 86.9%Improved
use INVERT instead of Network Dissectionneuron labeling · MaskRCNN ResNet50 FPN on MS COCOexplanation accuracy: 95.06% -> 98.77%Improved
use INVERT instead of Network Dissectionneuron labeling · FCN ResNet50 on MS COCOexplanation accuracy: 95.24% -> 95.24%No clear effect
increase concept-fraction constraint alpha and formula length L (ablation)INVERT · ImageNet feature-extractor neurons (ResNet18, GoogleNet, EfficientNet B0, ViT B 16)AUC of explanationMixed
add a Goedel AND over the top-L neurons instead of a single neuron (ablation)training-free transfer, EfficientNet B4 · Caltech101 46 classesaccuracy: 39.96% -> 86.11%Improved
add a Goedel AND over the top-L neurons (L=1 to 5) (ablation)training-free transfer, DenseNet161 · Caltech101 46 classesaccuracy: 62.50% -> 75.00%Improved
add a Goedel AND over the top-L neurons (L=1 to 10) (ablation)training-free transfer, ViT L 16 · Caltech101 46 classesaccuracy: 47.12% -> 79.49%Improved
add a Goedel AND over the top-L neurons (L=1 to 10) (ablation)training-free transfer, AlexNet · Caltech101 46 classesaccuracy: 43.91% -> 30.88%Hurt
use a single INVERT-selected neuron instead of a trained linear headtraining-free transfer, DenseNet161 · Caltech101 46 classesaccuracy: 97.76% -> 62.50%Hurt
compare Goedel against Product, Lukasiewicz and Yager fuzzy operators (ablation)neuron circuits · random compositional concepts on four ImageNet modelsmean AUCImproved

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