Research map / Graph learning / GNN architectures
Topological Pooling on Graphs
Yuzhou Chen, Yulia R. Gel
This paper tried a common change: Add attention module
Add the attention mechanism · COX2, PTC_MM, IMDB-B · accuracy: 87.24$\pm$3.15 -> 85.58$\pm$3.53; 76.76$\pm$5.78 -> 75.12$\pm$5.59; 78.40$\pm$1.50 -> 76.20$\pm$1.98 Improved
What 492 accepted papers report for this change
- 444 improved
- 20 no clear effect
- 4 mixed
- 24 hurt
- Multi-Scale Wavelet Transformers for Operator Learning of Dynamical Systems ICML 2026 · add wavelet attention block · CKF step 64 · Rel L2: 1.09 -> 0.35Improved
- DA-DFGAS:Differentiable Federated Graph Neural Architecture Search with Distribution-Aware Attentive Aggregation AAAI 2026 · add attention mechanism to aggregation · Table 4 datasets · parameters: about 5-10% more parametersHurt
- Caracal: Causal Architecture via Spectral Mixing ICML 2026 · add sliding window attention layers to MHF stack · Avg accuracy · avg accuracy (%): 48.22 -> 49.01Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add local topological pooling enhanced graph convolution (TPGCL) (ablation) | Wit-TopoPool · COX2, PTC_MM, IMDB-B | accuracy: 87.24$\pm$3.15 -> 82.67$\pm$3.26; 76.76$\pm$5.78 -> 67.38$\pm$5.33; 78.40$\pm$1.50 -> 73.93$\pm$1.83 | Improved |
| Add the witness-complex topological layer (ablation) | Wit-TopoPool · COX2, PTC_MM, IMDB-B | accuracy: 87.24$\pm$3.15 -> 85.21$\pm$3.20; 76.76$\pm$5.78 -> 70.58$\pm$5.29; 78.40$\pm$1.50 -> 77.00$\pm$1.69 | Improved |
| Add the attention mechanism (ablation) | Wit-TopoPool · COX2, PTC_MM, IMDB-B | accuracy: 87.24$\pm$3.15 -> 85.58$\pm$3.53; 76.76$\pm$5.78 -> 75.12$\pm$5.59; 78.40$\pm$1.50 -> 76.20$\pm$1.98 | Improved |
| Choose landmarks randomly rather than by centrality (ablation) | Wit-TopoPool · COX2, PTC_MM | accuracy: 87.24$\pm$3.15 -> 82.98$\pm$3.88; 76.76$\pm$5.78 -> 71.53$\pm$6.17 | Hurt |
| Use betweenness-centrality landmarks instead of degree-centrality landmarks (ablation) | Wit-TopoPool · COX2, PTC_MM | accuracy: 87.24$\pm$3.15 -> 85.10$\pm$2.52; 76.76$\pm$5.78 -> 79.12$\pm$4.45 | Mixed |
| Use the arctan-weighted persistence score instead of summed lifespans (ablation) | Wit-TopoPool · COX2, PTC_MM | accuracy: 87.24$\pm$3.15 -> 79.78$\pm$1.06; 79.12$\pm$4.45 -> 76.18$\pm$5.00 | Hurt |
| Replace the witness complex with a Vietoris-Rips complex (ablation) | Wit-TopoPool · COX2, PTC_MM, IMDB-B | accuracy: 87.24$\pm$3.15 vs 87.63$\pm$5.00; 76.76$\pm$5.78 vs 76.17$\pm$6.47; 78.40$\pm$1.50 vs 71.10$\pm$2.55 | Mixed |
| Combine local and global topological representations with node features | Wit-TopoPool · BZR, COX2, MUTAG, PROTEINS, PTC_MR, PTC_MM, PTC_FM, PTC_FR, IMDB-B, IMDB-M, REDDIT-B | accuracy | Improved |
Closest papers by meaning
- Boosting Graph Pooling with Persistent Homology (NeurIPS 2024 · similarity 0.92)
- TopInG: Topologically Interpretable Graph Learning via Persistent Rationale Filtration (ICML 2025 · similarity 0.90)
- Geometry-Aware Edge Pooling for Graph Neural Networks (NeurIPS 2025 · similarity 0.89)
- Weisfeiler and Lehman Go Paths: Learning Topological Features via Path Complexes (AAAI 2024 · similarity 0.89)
- TopER: Topological Embeddings in Graph Representation Learning (NeurIPS 2025 · similarity 0.89)
- Topological Blindspots: Understanding and Extending Topological Deep Learning Through the Lens of Expressivity (ICLR 2025 · similarity 0.89)
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.
