Trans-Encoder-RoBERTa-base-cross (unsup.)

Reported on 5 benchmarks across 1 task · 1 paper

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· 2021-09-27
0.7903
best: 0.8689 (AnglE-LLaMA-13B)
Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations arXiv:2109.13059
Semantic Textual SimilarityonSTS15
Spearman Correlation· 2021-09-27
0.8577
best: 0.9004 (PromptEOL+CSE+LLaMA-30B)
Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations arXiv:2109.13059
Semantic Textual SimilarityonSTS Benchmark
Spearman Correlation· 2021-09-27
0.8465
best: 0.931 (Mnet-Sim)
Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations arXiv:2109.13059
Semantic Textual SimilarityonSTS12
Spearman Correlation· 2021-09-27
0.7637
best: 0.802 (PromptEOL+CSE+OPT-13B)
Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations arXiv:2109.13059
Semantic Textual SimilarityonSTS16
Spearman Correlation· 2021-09-27
0.8377
best: 0.87 (AnglE-LLaMA-7B-v2)
Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations arXiv:2109.13059