List your company
ResearchFormerly WizWand

Word Similarity on SimLex999 (test)

0.789Spearman CorrelationAUXGAN
0.319960.441730.56350.68527Sep 11, 2018Oct 16, 2019Nov 19, 2020Dec 24, 2021Jan 28, 2023Mar 3, 2024Apr 8, 2025
Updated 8mo ago

Evaluation results

MethodLinks
AUXGANVector space=GLOVE-CC2018.090.789
POST-DFFNVector space=GLOVE-CC2018.090.785
AUXGANVector space=SGNS-W22018.090.782
ATTRACT-REPELVector space=GLOVE-CC2018.090.781
POST-DFFNVector space=SGNS-W22018.090.781
ATTRACT-REPELVector space=SGNS-W22018.090.778
POST-DFFNVector space=FASTTEXT2018.090.768
AUXGANVector space=FASTTEXT2018.090.766
ATTRACT-REPELVector space=FASTTEXT2018.090.761
SOTA (Multi)Type=Special-purpose SOTA model (Multi vector)2023.050.605
Flan-T5-SDBackbone=Flan-T5-xxl, Clusters (k)=52025.040.555
SOTA (Single)Type=Special-purpose SOTA model (Single vector)2023.050.554
ROBERTA2STATIC_sentParent Model=ROBERTA-12, Dim.=7682021.060.5507
BERT2STATIC_sentParent Model=BERT-12, Dim.=7682021.060.5461
GPT2STATIC_paraParent Model=GPT2-12, Dim.=7682021.060.5417
SimminSense Selection=Minimum Sense Cosine2023.050.54
GPT2-1.5BCategory=Large existing models2023.050.523
Sim12Sense Selection=Sense 12 (relatedness)2023.050.522
Sim14Sense Selection=Sense 14 (Verb objects, nmod nouns)2023.050.5
BERT-SDBackbone=BERT, Clusters (k)=52025.040.499
GPT-J-6BCategory=Large existing models2023.050.492
Trnsf 124MType=Baseline Transformer2023.050.478
DRG2VecType=Discrete embeddings2025.040.476
Dict2VecType=Discrete embeddings2025.040.444
word2vecEmbedding Type=Classic Non-Contextual2023.050.442
Distributional (X)Vector space=SGNS-W22018.090.414
Distributional (X)Vector space=GLOVE-CC2018.090.407
Distributional (X)Vector space=FASTTEXT2018.090.383
GloVeEmbedding Type=Classic Non-Contextual2023.050.371
Word2VecType=Discrete embeddings2025.040.338