AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Neural Interaction Transparency (NIT): Disentangling Learned Interactions for Improved Interpretability

Michael Tsang, Hanpeng Liu, Sanjay Purushotham, Pavankumar Murali, Yan Liu

NeurIPS 2018 · Probabilistic and scientific modeling › Molecules and proteins · approach: Probabilistic embeddings and relational inference · 38 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
raise max interaction order K from 2 to 4 (ablation)NIT · Bike SharingRMSE: 0.31 ± 0.013 -> 0.240 ± 0.0097Improved
raise max interaction order K from 2 to 4 (ablation)NIT · Cal HousingRMSE: 0.448 ± 0.0080 -> 0.43 ± 0.013Improved
raise max interaction order K from 2 to 6 (ablation)NIT · MIMIC-IIIAUC: 0.76 ± 0.011 -> 0.77 ± 0.011No clear effect
raise max interaction order K from 10 to 20 (ablation)NIT · CIFAR-10 binaryAUC: 0.849 ± 0.0049 -> 0.860 ± 0.0034Improved
allow pairwise interactions (NIT K=2) instead of univariate GAMGAM · Bike SharingRMSE: 0.55 ± 0.016 -> 0.31 ± 0.013Improved
disentangle interactions with NIT instead of plain L1/L2 regularized MLPfeedforward network · synthetic x1x2 + x3x4entangled neurons in second layerImproved
disentangle through all layers without network blocks (ablation)NIT · Cal HousingRMSE: 0.43 RMSE at max order 3No clear effect
disentangle through all layers without network blocks (ablation)NIT · Bike SharingRMSE: 0.26 RMSE at max order 6No clear effect

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