Definitions come from WordNet, a hand-curated dictionary.
the right to buy or sell property at an agreed price; the right is purchased and if it is not exercised by a stated date the money is forfeited
choix is French for options, and every measurement below was made on that English word. Definitions come from English WordNet; Open Multilingual WordNet supplies the French headword, not a French definition.
Panels 3, 4 and 5 are computed per model. Switch to see them disagree.
Definitions come from WordNet, a hand-curated dictionary.
This panel splits choix itself, not options — a tokenizer does not care what language it is fed. It is the only panel here that does.
This is the tokenizer splitting text, not the model understanding it.
Shade shows how close the model puts each word to options, 1 of 8 also appear in the dictionary.
Shade shows how close the model puts each word to options, 2 of 8 also appear in the dictionary.
Shade shows how close the model puts each word to options, 1 of 8 also appear in the dictionary.
Model neighbours are distributional, not dictionary synonyms. Two words can be close because they appear in similar sentences, which is why an antonym can outscore a synonym here.
Scores are projections onto axes we defined from anchor words, not labels the model assigns.
These sit just outside the closest neighbours in panel 3, and no dictionary lists any of them as related to options. That gap is the model’s own learned association.
These sit just outside the closest neighbours in panel 3, and no dictionary lists any of them as related to options. That gap is the model’s own learned association.
These sit just outside the closest neighbours in panel 3, and no dictionary lists any of them as related to options. That gap is the model’s own learned association.
These are statistical associations in the training data, not the model thinking.