Research map / Graph learning / Graph self-supervised learning
Does Invariant Graph Learning via Environment Augmentation Learn Invariance?
Yongqiang Chen, Yatao Bian, Kaiwen Zhou, Binghui Xie, Bo Han, James Cheng
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use an environment assistant to split graphs by whether its proxy predictions are correct, then maximize cross-set intra-class subgraph mutual information | CIGA-style invariant graph learning · Two-piece graphs, {0.7,0.9} | OOD accuracy: 43.62±3.20 -> 72.50±1.06 | Improved |
| Use an environment assistant and cross-set contrastive sampling | CIGA-style invariant graph learning · Two-piece average | OOD accuracy: 67.18 -> 79.21 | Improved |
| Use an environment assistant and cross-set contrastive sampling | CIGA-style invariant graph learning · EC50-Assay, EC50-Scaffold, EC50-Size, Ki-Assay, Ki-Scaffold, Ki-Size, CMNIST-sp, Graph-SST2 | OOD accuracy: 76.42±1.59 -> 77.56±2.88; 64.56±1.25 -> 66.28±0.45; 61.61±1.52 -> 64.25±1.21; 74.61±2.28 -> 77.92±2.48; 69.38±1.65 -> 73 | Improved |
| Add the CIGAv2 constraint to GALA (ablation) | GALA · Ki-Scaffold | OOD performance | Improved |
| Add the CIGAv2 constraint to GALA (ablation) | GALA · CMNIST-sp | OOD performance | No clear effect |
| Set the contrastive penalty weight to zero (ablation) | GALA · Two-piece {0.7,0.9} | OOD performance | Hurt |
| Set the positive/negative group upsampling times to zero (ablation) | GALA · Two-piece {0.7,0.9} | OOD performance | Hurt |
| Vary the contrastive penalty weight or upsampling times away from zero (ablation) | GALA · Two-piece {0.7,0.9} | OOD performance | No clear effect |
| Infer environment labels or generate virtual environments without ensuring faithful environment information | Invariant graph learning methods · Two-piece graphs | OOD performance | Hurt |
| Use GALA rather than CIGA | Invariant graph learning · CMNIST-sp | OOD accuracy: 15.83±2.56 -> 68.94±0.56 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DisC — DiSC: Differential Spectral Clustering of Features (NeurIPS 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Graph Invariant Learning with Subgraph Co-mixup for Out-of-Distribution Generalization (AAAI 2024 · similarity 0.91)
- Dissecting the Failure of Invariant Learning on Graphs (NeurIPS 2024 · similarity 0.90)
- Enhancing Graph Invariant Learning from a Negative Inference Perspective (ICML 2025 · similarity 0.90)
- Disentangling Invariant Subgraph via Variance Contrastive Estimation under Distribution Shifts (ICML 2025 · similarity 0.89)
- Improving Out-of-Distribution Generalization in Graphs via Hierarchical Semantic Environments (CVPR 2024 · similarity 0.89)
- Empowering Graph Invariance Learning with Deep Spurious Infomax (ICML 2024 · similarity 0.88)
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.
