AI Paper Atlas

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

Feature Interaction Interpretability: A Case for Explaining Ad-Recommendation Systems via Neural Interaction Detection

Michael Tsang, Dehua Cheng, Hanpeng Liu, Xue Feng, Eric Zhou, Yan Liu

ICLR 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Self-explaining and visual explanation networks · 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
Encode AutoInt's top-40 detected interactions in DeepFMDeepFM · CriteoAUC: 0.8079 ±3e-4 -> 0.8097± 2e-4Improved
Encode AutoInt's top-40 detected interactions in DeepFMDeepFM · Criteologloss: 0.4436±2e-4 -> 0.4420±2e-4Improved
Encode AutoInt's top-10 detected interactions in DeepFMDeepFM · AvazuAUC: 0.7792±3e-4 -> 0.7795±2e-4Improved
Encode AutoInt's top-10 detected interactions in DeepFMDeepFM · Avazulogloss: 0.3804± 9e-5 -> 0.3802±2e-4Improved
Encode AutoInt's top-40 detected interactions in AutoIntAutoInt · CriteoAUC: 0.8083 -> 0.8090±{2e-4}Improved
Encode AutoInt's top-40 detected interactions in AutoIntAutoInt · Criteologloss: 0.4434 -> 0.4426±{2e-4}Improved
Encode AutoInt's top-10 detected interactions in AutoIntAutoInt · AvazuAUC: 0.7774 -> 0.7773±1e-4No clear effect
Encode AutoInt's top-10 detected interactions in AutoIntAutoInt · Avazulogloss: 0.3811 -> 0.3811±5e-5No clear effect
Add detected feature interactions up to rank threshold LMADEX explanations · DNA-CNN test instancesmean-squared error: 10e-3±1e-3 -> 5.0e-3±8e-4Improved
Add detected feature interactions up to rank threshold LMADEX explanations · Sentiment-LSTM test instancesmean-squared error: 8.0e-2±6e-3 -> 0.4e-2±3e-3Improved
Add detected feature interactions up to rank threshold LMADEX explanations · ResNet152 test instancesmean-squared error: 1.9±0.1 -> 0.9±0.2Improved
Add detected node interactions up to rank threshold LMADEX explanations · GCN Cora test instancesmean-squared error: 7.1e3±7e2 -> 2e3±1e3Improved

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