Research map / Graph learning / GNN architectures
Differentiable Lifting for Topological Neural Networks
Jorge Luiz Franco, Gabriel Duarte, Alexander V Nikitin, Moacir A Ponti, Diego Mesquita, Amauri H Souza
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 |
|---|---|---|---|
| replace cycle lifting with differentiable lifting | CWN · NCI1 | accuracy: 76.93 -> 79.81 | Improved |
| replace cycle lifting with differentiable lifting | CWN · MUTAG | accuracy: 66.67 -> 85.96 | Improved |
| replace cycle lifting with differentiable lifting | CWN · ZINC | MAE: 0.46 -> 0.17 | Improved |
| replace cycle lifting with differentiable lifting | CXN · NCI1 | accuracy: 72.02 -> 82.08 | Improved |
| replace cycle lifting with differentiable lifting | CXN · Proteins | accuracy: 70.83 -> 69.94 | Hurt |
| replace cycle lifting with differentiable lifting | CIN · MUTAG | accuracy: 66.96 -> 88.72 | Improved |
| replace k-hop lifting with differentiable lifting | UniGCN2 · MOLHIV | accuracy: 50.72 -> 69.32 | Improved |
| replace k-NN lifting with differentiable lifting | UniGIN · NCI1 | accuracy: 72.83 -> 64.88 | Hurt |
| replace GIN backbone with GPS in dLift (ablation) | CWN · ZINC | MAE: 0.46 -> 0.17 | Improved |
| replace GIN backbone with GPS in dLift (ablation) | CWN · MOLHIV | accuracy: 75.37 -> 64.31 | Hurt |
| replace cycle lifting with differentiable lifting | CWN · Node classification average | accuracy: 73.52 -> 77.74 | Improved |
| replace cycle lifting with differentiable lifting | TopoTune · Node classification average | accuracy: 70.91 -> 75.84 | Improved |
Closest papers by meaning
- Topological Blindspots: Understanding and Extending Topological Deep Learning Through the Lens of Expressivity (ICLR 2025 · similarity 0.87)
- Weisfeiler and Lehman Go Paths: Learning Topological Features via Path Complexes (AAAI 2024 · similarity 0.86)
- Diffusing to the Top: Boost Graph Neural Networks with Minimal Hyperparameter Tuning (ICLR 2025 · similarity 0.85)
- TopoTune: A Framework for Generalized Combinatorial Complex Neural Networks (ICML 2025 · similarity 0.85)
- A Topological Perspective on Demystifying GNN-Based Link Prediction Performance (ICLR 2024 · similarity 0.85)
- TopInG: Topologically Interpretable Graph Learning via Persistent Rationale Filtration (ICML 2025 · similarity 0.84)
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.
