Research map / Vision-language and multimodal / Cross-modal learning
Rank-Aware Hyperbolic Alignment for Vision–Language Dataset Distillation
Jongoh Jeong, Sun-Kyung Lee, KUK-JIN YOON
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 |
|---|---|---|---|
| use RAHA instead of CovMatch at larger budgets | RAHA · Flickr8k | mean retrieval recall: 25.3 vs 18.8 (200 pairs); 30.7 vs 25.9 (500) | Improved |
| use RAHA instead of CovMatch at 100 pairs | RAHA · Flickr30k | mean retrieval recall: 22.8 -> 20.7 | Hurt |
| use RAHA instead of LoRS at 100 pairs | RAHA · Flickr30k | mean retrieval recall: 27.4 -> 20.7 | Hurt |
| use RAHA instead of CovMatch | RAHA · Flickr8k cross-architecture | mean recall: 7.3 -> 6.9 (100); 7.2 -> 8.7 (200); 8.7 -> 12.7 (500) | Mixed |
| use RAHA instead of CovMatch | RAHA · Flickr8k image-side noise | mean degradation: 4.6 -> 4.1 (100); 6.4 -> 6.8 (500) | Mixed |
| add range relevance term to hyperbolic contrast (ablation) | RAHA · Flickr8k 100 pairs | mean retrieval | Improved |
| use residual term alone (ablation) | RAHA · Flickr8k 100 pairs | mean retrieval | Hurt |
| combine range and residual with regularizers (ablation) | RAHA · Flickr8k 100 pairs | mean retrieval | Improved |
| raise energy threshold rho from 0.95 to 1.0 (ablation) | RAHA · retrieval | mean retrieval | Hurt |
| use larger real batches (ablation) | RAHA · retrieval | mean retrieval | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CovMatch — CovMatch: Cross-Covariance Guided Multimodal Dataset Distillation with Trainable Text Encoder (NeurIPS 2025) · better on 8 of 12 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- HyperET: Efficient Training in Hyperbolic Space for Multi-modal Large Language Models (NeurIPS 2025 · similarity 0.90)
- DH-Set: Improving Vision-Language Alignment with Diverse and Hybrid Set-Embeddings Learning (CVPR 2025 · similarity 0.90)
- Compositional Entailment Learning for Hyperbolic Vision-Language Models (ICLR 2025 · similarity 0.88)
- Locality Alignment Improves Vision-Language Models (ICLR 2025 · similarity 0.88)
- Assessing and Learning Alignment of Unimodal Vision and Language Models (CVPR 2025 · similarity 0.88)
- Topology-Aware Representation Alignment for Semi-Supervised Vision-Language Learning (NeurIPS 2026 · 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.
