1
Your word
measured
“for your own good”
Definitions come from WordNet, a hand-curated dictionary.
benefit
利益 is Japanese for good, 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.
“for your own good”
Definitions come from WordNet, a hand-curated dictionary.
2 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.
This panel splits 利益 itself, not good — 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.
bad is the dictionary opposite of good, yet this model puts it closer than every one of the 25 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 good, 1 of 8 also appear in the dictionary. 25 of 36 dictionary synonyms are in this build; the rest have no vector to compare yet.
bad is the dictionary opposite of good, yet this model puts it closer than 24 of the 25 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 good. 25 of 36 dictionary synonyms are in this build; the rest have no vector to compare yet.
bad is the dictionary opposite of good, yet this model puts it closer than every one of the 25 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 good. 25 of 36 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 good. 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 good. 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 good. That gap is the model’s own learned association.
These are statistical associations in the training data, not the model thinking.