SimCSE-RoBERTalarge

Reported on 5 benchmarks across 1 task · 1 paper · 4 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing5 results

Semantic Textual SimilarityonSTS14
Spearman Correlation· 2021-04-18
0.8236
best: 0.8689 (AnglE-LLaMA-13B)
SOTA
SimCSE: Simple Contrastive Learning of Sentence Embeddings arXiv:2104.08821
Semantic Textual SimilarityonSTS15
Spearman Correlation· 2021-04-18
0.8666
best: 0.9004 (PromptEOL+CSE+LLaMA-30B)
SOTA
SimCSE: Simple Contrastive Learning of Sentence Embeddings arXiv:2104.08821
Semantic Textual SimilarityonSICK
Spearman Correlation· 2021-04-18
0.8195
best: 0.8243 (PromCSE-RoBERTa-large (0.355B))
SOTA
SimCSE: Simple Contrastive Learning of Sentence Embeddings arXiv:2104.08821
Semantic Textual SimilarityonSTS16
Spearman Correlation· 2021-04-18
0.8393
best: 0.87 (AnglE-LLaMA-7B-v2)
SOTA
SimCSE: Simple Contrastive Learning of Sentence Embeddings arXiv:2104.08821
Semantic Textual SimilarityonSTS Benchmark
Spearman Correlation· 2021-04-18
0.867
best: 0.931 (Mnet-Sim)
SimCSE: Simple Contrastive Learning of Sentence Embeddings arXiv:2104.08821