Research map / Label-efficient and robust learning / Recognition and long-tail
Unsupervised Hyperbolic Representation Learning via Message Passing Auto-Encoders
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 HGCAE-H instead of DBGAN | Cora link prediction · Cora | AUC: 0.945 -> 0.956 | Helped |
| Use HGCAE-H instead of DBGAN | Citeseer link prediction · Citeseer | AUC: 0.945 -> 0.967 | Helped |
| Use HGCAE-H instead of GALA | BlogCatalog node clustering · BlogCatalog | ACC: 0.400 -> 0.741 | Helped |
| Use HGCAE-H instead of GALA | Amazon Photo node clustering · Amazon Photo | ACC: 0.512 -> 0.817 | Helped |
| Use HGCAE-P instead of GALA | Wiki node clustering · Wiki | ACC: 0.544 -> 0.459 | Hurt |
| Use HGCAE-P instead of P-VAE | Phylogenetic link prediction · Phylogenetic | AUC: 0.590 -> 0.688 | Helped |
| Use HGCAE-P instead of P-VAE | CS PhDs link prediction · CS PhDs | AUC: 0.598 -> 0.673 | Helped |
| Use HGCAE-P instead of P-VAE | Diseases link prediction · Diseases | AUC: 0.923 -> 0.926 | Helped |
| Add node-attribute reconstruction alongside graph reconstruction (ablation) | GAE link prediction · Cora | AUC: 91.0 -> 92.7 | Helped |
| Add fixed-curvature hyperbolic message passing (ablation) | Ablation I link prediction model · Cora | AUC: 92.7 -> 94.6 | Helped |
| Add geometry-aware attention and fixed-curvature hyperbolic message passing (ablation) | GAE link prediction · Citeseer | AUC: 89.5 -> 96.1 | Helped |
| Add reconstruction of both graph structure and node attributes to the hyperbolic attention model (ablation) | Ablation III link prediction model · Citeseer | AUC: 96.1 -> 96.7 | Helped |
Closest papers by meaning
- Hyperbolic Representation Learning: Revisiting and Advancing (ICML 2023 · similarity 0.91)
- Hyperbolic Embeddings of Supervised Models (NeurIPS 2024 · similarity 0.90)
- Fully Hyperbolic Convolutional Neural Networks for Computer Vision (ICLR 2024 · similarity 0.87)
- Learning Structured Representations with Hyperbolic Embeddings (NeurIPS 2024 · similarity 0.87)
- Weighted Embeddings for Low-Dimensional Graph Representation (AAAI 2025 · similarity 0.87)
- Learning Along the Arrow of Time: Hyperbolic Geometry for Backward-Compatible Representation Learning (ICML 2025 · similarity 0.86)
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.