Research map / Label-efficient and robust learning / Recognition and long-tail
Poincaré Embeddings for Learning Hierarchical Representations
Read the paperCheck 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 Poincaré rather than Euclidean embeddings | WordNet noun hierarchy reconstruction · WordNet noun hierarchy | mean rank: 1157.3 -> 4.9 | Helped |
| Use Poincaré rather than Translational embeddings | WordNet noun hierarchy reconstruction · WordNet noun hierarchy | MAP: 0.565 -> 0.823 | Helped |
| Use Poincaré rather than Euclidean embeddings | WordNet noun hierarchy link prediction · WordNet noun hierarchy | MAP: 0.490 -> 0.825 | Helped |
| Use Poincaré rather than Translational embeddings | WordNet noun hierarchy link prediction · WordNet noun hierarchy | MAP: 0.562 -> 0.825 | Helped |
| Use Poincaré rather than Euclidean embeddings | Network link prediction · AstroPh, CondMat, GrQc, and HepPh | MAP: AstroPh: 0.508 -> 0.671; CondMat: 0.308 -> 0.539; GrQc: 0.438 -> 0.660; HepPh: 0.642 -> 0.683 | Helped |
| Use Poincaré rather than Euclidean embeddings | Network reconstruction · AstroPh, CondMat, GrQc, and HepPh | MAP: AstroPh: 0.376 -> 0.703; CondMat: 0.356 -> 0.799; GrQc: 0.522 -> 0.990; HepPh: 0.434 -> 0.811 | Helped |
| Use Poincaré rather than Euclidean embeddings | HyperLex lexical entailment scoring · HyperLex noun pairs | Spearman's ρ: 0.389 -> 0.512 | Helped |
| Use WordNet-trained Poincaré embeddings | Non-graded lexical entailment · Wbless | accuracy: 0.86 | Helped |
Closest papers by meaning
- Weighted Embeddings for Low-Dimensional Graph Representation (AAAI 2025 · similarity 0.90)
- Learning Continuous Hierarchies in the Lorentz Model of Hyperbolic Geometry (ICML 2018 · similarity 0.89)
- Hyperbolic Embeddings of Supervised Models (NeurIPS 2024 · similarity 0.88)
- Learning Mixed-Curvature Representations in Product Spaces (ICLR 2019 · similarity 0.88)
- Multi-relational Poincaré Graph Embeddings (NeurIPS 2019 · similarity 0.88)
- The Numerical Stability of Hyperbolic Representation Learning (ICML 2023 · 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.