AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

Learning Substructure Invariance for Out-of-Distribution Molecular Representations

Nianzu Yang, Kaipeng Zeng, Qitian Wu, Xiaosong Jia, Junchi Yan

NeurIPS 2022 · Label-efficient and robust learning › OOD and anomaly detection · approach: Robust fine-tuning for OOD generalization · 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
add attention-based substructure architecture with ERM loss (ablation)GIN + ERM · EC50-assayROC-AUC (%): 69.35 -> 71.44Improved
add attention-based substructure architecture with ERM loss (ablation)GIN + ERM · EC50-sizeROC-AUC (%): 60.94 -> 64.23Improved
add new learning objective on top of architecture (full model) (ablation)Our architecture + ERM loss · EC50-assayROC-AUC (%): 71.44 -> 73.25Improved
add new learning objective on top of architecture (full model) (ablation)Our architecture + ERM loss · EC50-sizeROC-AUC (%): 64.23 -> 65.09Improved
add substructure architecture to GIN + new objective (ablation)GIN + new learning objective · EC50-assayROC-AUC (%): 72.07 -> 73.25Improved
replace given environment labels with inferred environments (ablation)Our model · EC50-assayROC-AUC (%): 71.94 -> 73.25Improved
replace given environment labels with inferred environments (ablation)Our model · EC50-sizeROC-AUC (%): 63.38 -> 65.09Improved
use inferred environment labels instead of given ones (ablation)DANN · EC50-assayROC-AUC (%): 66.97 -> 68.83Improved
use inferred environment labels instead of given ones (ablation)DANN · EC50-sizeROC-AUC (%): 61.11 -> 62.56Improved
adapt backbone with our methodGCN · BACEROC-AUC (%): 80.01 -> 84.33Improved
adapt backbone with our methodGIN · HIVROC-AUC (%): 76.58 -> 78.31Improved
add virtual node to GCNGCN · BACEROC-AUC (%): 80.01 -> 77.51Hurt

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