1
Your word
measured
“he collected dry sticks for a campfire”
Definitions come from WordNet, a hand-curated dictionary.
an implement consisting of a length of wood
棒 is Japanese for stick, and every measurement below was made on that English word. Definitions come from English WordNet; Open Multilingual WordNet supplies the Japanese headword, not a Japanese definition.
Panels 3, 4 and 5 are computed per model. Switch to see them disagree.
“he collected dry sticks for a campfire”
Definitions come from WordNet, a hand-curated dictionary.
3 of those are a raw byte rather than a character: Qwen3 8B splits this word below the character, so no single token it sees is readable.
3 of those are a raw byte rather than a character: Mistral 7B splits this word below the character, so no single token it sees is readable.
This panel splits 棒 itself, not stick — 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.
move is the dictionary opposite of stick, yet this model puts it closer than 13 of the 17 synonyms it can score. Opposites share the sentences a word lives in, so cosine alone cannot tell them apart - which is why the dictionary seeds this list rather than the geometry.
Shade shows how close the model puts each word to stick. 17 of 40 dictionary synonyms are in this build; the rest have no vector to compare yet.
move is the dictionary opposite of stick, yet this model puts it closer than 8 of the 17 synonyms it can score. Opposites share the sentences a word lives in, so cosine alone cannot tell them apart - which is why the dictionary seeds this list rather than the geometry.
Shade shows how close the model puts each word to stick, 1 of 8 also appear in the dictionary. 17 of 40 dictionary synonyms are in this build; the rest have no vector to compare yet.
move is the dictionary opposite of stick, yet this model puts it closer than 12 of the 17 synonyms it can score. Opposites share the sentences a word lives in, so cosine alone cannot tell them apart - which is why the dictionary seeds this list rather than the geometry.
Shade shows how close the model puts each word to stick. 17 of 40 dictionary synonyms are in this build; the rest have no vector to compare yet.
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 stick. 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 stick. 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 stick. That gap is the model’s own learned association.
These are statistical associations in the training data, not the model thinking.