Semantic Textual Similarity On Cxc
Metrics
avg ± std
Results
Performance results of various models on this benchmark
| Paper Title | ||
|---|---|---|
| PromCSE-RoBERTa-large (0.355B) | 74.8± 1.0 | Improved Universal Sentence Embeddings with Prompt-based Contrastive Learning and Energy-based Learning |
| DE-T2T+I2T | 74.5 ± 0.4 | MURAL: Multimodal, Multitask Retrieval Across Languages |
| MURAL-large | 74.1 ± 0.4 | MURAL: Multimodal, Multitask Retrieval Across Languages |
| ALIGN-L2 | 72.9 ± 0.4 | MURAL: Multimodal, Multitask Retrieval Across Languages |
0 of 4 row(s) selected.