Research map / Vision-language and multimodal / Cross-modal learning
Cross-Linked Unified Embedding for cross-modality representation learning
Xinming Tu, Zhi-Jie Cao, Chen-Rui Xia, Sara Mostafavi, Ge Gao
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 |
|---|---|---|---|
| add adversarial modality discriminator loss (ablation) | CLUE · NeurIPS 2021 multimodal competition test | matching score Overall: 0.0284 -> 0.0539 | Improved |
| add mean squared error alignment loss (ablation) | CLUE · NeurIPS 2021 multimodal competition test | matching score Overall: 0.0451 -> 0.0539 | Improved |
| add cross-linked loss (cross-encoders) (ablation) | CLUE · NeurIPS 2021 multimodal competition test | matching score Overall: 0.0444 -> 0.0539 | Improved |
| use all three loss components versus none (ablation) | CLUE · NeurIPS 2021 multimodal competition test | matching score Overall: 0.0001 -> 0.0539 | Improved |
| CLUE versus Cobolt | CLUE · Competition data benchmark | matching score Overall: 0.0244 -> 0.0539 | Improved |
| CLUE versus MultiVI | CLUE · Competition data benchmark | matching score Overall: 0.0169 -> 0.0539 | Improved |
| CLUE versus Bridge-integration | CLUE · Competition data benchmark | matching score Overall: 0.0116 -> 0.0539 | Improved |
| CLUE versus MultiVI (second best) | CLUE · SHARE-seq | matching score Overall: 0.0601 -> 0.135 | Improved |
| CLUE versus Bridge-integration | CLUE · SHARE-seq | matching score Overall: 0.0335 -> 0.135 | Improved |
| CLUE versus Bridge-integration (best baseline) | CLUE · Competition data benchmark | FOSCTTM Overall (lower better): 0.0344 -> 0.0162 | Improved |
| CLUE versus MultiVI | CLUE · Competition data benchmark | FOSCTTM Overall (lower better): 0.0635 -> 0.0162 | Improved |
| CLUE versus Top2 competition method | CLUE · NeurIPS 2021 multimodal competition test | matching score Overall: 0.0427 -> 0.0539 | Improved |
Closest papers by meaning
- A Cross Modal Knowledge Distillation & Data Augmentation Recipe for Improving Transcriptomics Representations through Morphological Features (ICML 2025 · similarity 0.87)
- Multi-modal Transfer Learning between Biological Foundation Models (NeurIPS 2024 · similarity 0.86)
- An Information Criterion for Controlled Disentanglement of Multimodal Data (ICLR 2025 · similarity 0.86)
- Connect, Collapse, Corrupt: Learning Cross-Modal Tasks with Uni-Modal Data (ICLR 2024 · similarity 0.85)
- ES-Merging: Biological MLLM Merging via Embedding Space Signals (NeurIPS 2026 · similarity 0.85)
- RePercENT: Scaling Disentangled Representation Learning Beyond Two Modalities (NeurIPS 2026 · similarity 0.85)
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.
