Title | Type | Venue | Code | Year | State | Date | |
---|---|---|---|---|---|---|---|
0 | Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training data | ⚖Stability | 📝arXiv'2021 | 2021 | Removed | 2023-11-07 | |
1 | (Provable) Adversarial Robustness for Group Equivariant Tasks: Graphs, Point Clouds, Molecules, and More | 🔐Certification | 📝NeurIPS'2023 | Code | 2023 | Added | 2023-11-07 |
2 | Hierarchical Randomized Smoothing | 🔐Certification | 📝NeurIPS'2023 | Code | 2023 | Added | 2023-11-07 |
3 | Adversarial Training for Graph Neural Networks: Pitfalls, Solutions, and New Directions | 🛡Defense | 📝NeurIPS | Code | 2023 | Added | 2023-11-07 |
4 | Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training data | ⚖Stability | 📝NeurIPS'2021 | Code | 2021 | Added | 2023-11-07 |