Word Similarity on WS-353 SIM (test)
0.807Spearman CorrelationSOTA (Multi)
Evaluation Results
| Method | Links | |
|---|---|---|
| SOTA (Multi)Type=Special-purpose SOTA model (Multi vector)2023.05 | 0.807 | |
| Flan-T5-SDBackbone=Flan-T5-xxl, Clusters (k)=52025.04 | 0.799 | |
| BERT-SDBackbone=BERT, Clusters (k)=52025.04 | 0.794 | |
| SOTA (Single)Type=Special-purpose SOTA model (Single vector)2023.05 | 0.764 | |
| Sim12Sense Selection=Sense 12 (relatedness)2023.05 | 0.749 | |
| DRG2VecType=Discrete embeddings2025.04 | 0.728 | |
| GPT2-1.5BCategory=Large existing models2023.05 | 0.706 | |
| Dict2VecType=Discrete embeddings2025.04 | 0.696 | |
| word2vecEmbedding Type=Classic Non-Contextual2023.05 | 0.684 | |
| Trnsf 124MType=Baseline Transformer2023.05 | 0.681 | |
| Word2VecType=Discrete embeddings2025.04 | 0.679 | |
| GPT-J-6BCategory=Large existing models2023.05 | 0.673 | |
| Sim14Sense Selection=Sense 14 (Verb objects, nmod nouns)2023.05 | 0.655 | |
| GloVeEmbedding Type=Classic Non-Contextual2023.05 | 0.607 | |
| SimminSense Selection=Minimum Sense Cosine2023.05 | 0.607 |