Semantic Textual Similarity On Sts Benchmark
Metrics
Spearman Correlation
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | Spearman Correlation |
|---|---|
| exploring-the-limits-of-transfer-learning | 0.886 |
| sentence-bert-sentence-embeddings-using | 0.8615 |
| universal-sentence-encoder | - |
| q8bert-quantized-8bit-bert | - |
| distilbert-a-distilled-version-of-bert | - |
| roberta-a-robustly-optimized-bert-pretraining | - |
| sentence-bert-sentence-embeddings-using | 0.7703 |
| structbert-incorporating-language-structures | 0.924 |
| generating-datasets-with-pretrained-language | 0.7782 |
| sentence-bert-sentence-embeddings-using | 0.7777 |
| ernie-20-a-continual-pre-training-framework | - |
| albert-a-lite-bert-for-self-supervised | - |
| smart-robust-and-efficient-fine-tuning-for | - |
| angle-optimized-text-embeddings | 0.8969 |
| trans-encoder-unsupervised-sentence-pair | 0.867 |
| ernie-enhanced-language-representation-with | - |
| sentence-bert-sentence-embeddings-using | 0.79 |
| angle-optimized-text-embeddings | 0.8897 |
| trans-encoder-unsupervised-sentence-pair | 0.839 |
| fast-effective-and-self-supervised | 0.764 |
| exploring-the-limits-of-transfer-learning | 0.85 |
| deep-continuous-prompt-for-contrastive-1 | 0.8787 |
| exploring-the-limits-of-transfer-learning | 0.921 |
| scaling-sentence-embeddings-with-large | 0.8914 |
| an-unsupervised-sentence-embedding-method | 0.6921 |
| generating-datasets-with-pretrained-language | 0.7651 |
| Model 27 | 0.7981 |
| big-bird-transformers-for-longer-sequences | .878 |
| 190910351 | - |
| exploring-the-limits-of-transfer-learning | - |
| llm-int8-8-bit-matrix-multiplication-for | - |
| adversarial-self-attention-for-language | 0.892 |
| def2vec-extensible-word-embeddings-from | 0.6372 |
| informer-transformer-likes-informed-attention | 0.8988 |
| fast-effective-and-self-supervised | 0.787 |
| mnet-sim-a-multi-layered-semantic-similarity-1 | 0.931 |
| spanbert-improving-pre-training-by | - |
| trans-encoder-unsupervised-sentence-pair | 0.8616 |
| exploring-the-limits-of-transfer-learning | - |
| exploring-the-limits-of-transfer-learning | 0.898 |
| adversarial-self-attention-for-language | 0.865 |
| sentence-bert-sentence-embeddings-using | 0.8479 |
| fnet-mixing-tokens-with-fourier-transforms | 0.84 |
| rematch-robust-and-efficient-matching-of | 0.6652 |
| q-bert-hessian-based-ultra-low-precision | - |
| scaling-sentence-embeddings-with-large | 0.8856 |
| bert-pre-training-of-deep-bidirectional | 0.865 |
| smart-robust-and-efficient-fine-tuning-for | 0.925 |
| smart-robust-and-efficient-fine-tuning-for | - |
| clear-contrastive-learning-for-sentence | - |
| a-statistical-framework-for-low-bitwidth | - |
| Model 52 | - |
| xlnet-generalized-autoregressive-pretraining | - |
| angle-optimized-text-embeddings | 0.8897 |
| simcse-simple-contrastive-learning-of | 0.867 |
| trans-encoder-unsupervised-sentence-pair | 0.8655 |
| charformer-fast-character-transformers-via | - |
| sentence-bert-sentence-embeddings-using | 0.8445 |
| scaling-sentence-embeddings-with-large | 0.8833 |
| Model 60 | - |
| on-the-sentence-embeddings-from-pre-trained | 0.7226 |
| trans-encoder-unsupervised-sentence-pair | 0.8465 |
| ernie-20-a-continual-pre-training-framework | - |
| how-to-train-bert-with-an-academic-budget | - |
| deberta-decoding-enhanced-bert-with | - |
| entailment-as-few-shot-learner | - |