AI Paper Atlas

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

Understanding Learned Models by Identifying Important Features at the Right Resolution

Kyubin Lee, Akshay Sood, Mark Craven

AAAI 2019 · Label-efficient and robust learning › Recognition and long-tail · approach: Feature attribution methods · 7 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
Increase the synthetic test-set size from m=32 to m=16384Proposed feature and interaction testing method · Synthetic datafeature power: 0.722 -> 0.975Improved
Increase the synthetic test-set size from m=32 to m=16384Proposed feature and interaction testing method · Synthetic datainteraction power: 0.132 -> 0.949Improved
Increase model-output noise from σ=0.00 to σ=1.28Proposed feature and interaction testing method · Synthetic datafeature power: 0.999 -> 0.770Hurt
Increase model-output noise from σ=0.00 to σ=1.28Proposed feature and interaction testing method · Synthetic datainteraction power: 0.991 -> 0.564Hurt
Increase synthetic test-set size across the evaluated valuesProposed feature and interaction testing method · Synthetic datafeature FDR: 0.019 at m=32; 0.032 at m=16384No clear effect
Increase model-output noise across the evaluated valuesProposed feature and interaction testing method · Synthetic datainteraction FDR: 0.000 at σ=0.00; 0.050 at σ=1.28No clear effect
Test interactions only after constituent features are identified as importantProposed interaction analysis · Synthetic datapowerHurt

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