causal

adjective 1 sense

involving or constituting a cause; causing

causal is French for causal, 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.

Story Show the data
English Español Français 日本語 中文
Qwen3 8B
DeepSeek V3
Mistral 7B

Panels 3, 4 and 5 are computed per model. Switch to see them disagree.

1 Your word measured

“a causal relationship between scarcity and higher prices”

adjective 1 sense 0 synonyms

Definitions come from WordNet, a hand-curated dictionary.

2 To an AI, it's pieces measured
Qwen3 8B causal 1 token
DeepSeek V3 causal 1 token
Mistral 7B caus al 2 tokens

This panel splits causal itself, not causal — 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.

3 Synonyms, antonyms, and the AI measured
Qwen3 8B
DeepSeek V3
Mistral 7B

The model’s neighbours

temporal0.12 sexual0.12 analytic0.12 moral0.12 dynamical0.12 conditional0.11 casual0.11 marginal0.11

Dictionary synonyms measured

WordNet lists none for this word.

Shade shows how close the model puts each word to causal.

The model’s neighbours

casual0.30 causes0.18 cause0.18 caused0.17 temporal0.17 explanatory0.16 causing0.16 factual0.16

Dictionary synonyms measured

WordNet lists none for this word.

Shade shows how close the model puts each word to causal.

This model splits causal into 2 pieces, so it has no vector of its own here: this is the average of its fragments, and the results below are correspondingly rough.

The model’s neighbours

paralegal0.37 quantal0.33 coral0.33 disposal0.33 postal0.32 basal0.32 reversal0.32 viral0.32

Dictionary synonyms measured

WordNet lists none for this word.

Shade shows how close the model puts each word to causal.

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.

4 Its personality measured
Qwen3 8B
DeepSeek V3
Mistral 7B
rare common
concrete abstract
casual formal
everyday technical
negative positive
mild intense
powerless powerful
small big
rare common
concrete abstract
casual formal
everyday technical
negative positive
mild intense
powerless powerful
small big

This model splits causal into 2 pieces, so it has no vector of its own here: this is the average of its fragments, and the results below are correspondingly rough.

rare common
concrete abstract
casual formal
everyday technical
negative positive
mild intense
powerless powerful
small big

Scores are projections onto axes we defined from anchor words, not labels the model assigns.

5 Surprising neighbours measured
Qwen3 8B
DeepSeek V3
Mistral 7B
contextual0.112 gaussian0.110 hypotheses0.109 seasonal0.108 spectral0.108

These sit just outside the closest neighbours in panel 3, and no dictionary lists any of them as related to causal. That gap is the model’s own learned association.

predictive0.152 moral0.152 modal0.142 spatial0.135 conceptual0.134

These sit just outside the closest neighbours in panel 3, and no dictionary lists any of them as related to causal. That gap is the model’s own learned association.

This model splits causal into 2 pieces, so it has no vector of its own here: this is the average of its fragments, and the results below are correspondingly rough.

costal0.314 astragalus0.311 nasal0.310 accidental0.309 parental0.309

These sit just outside the closest neighbours in panel 3, and no dictionary lists any of them as related to causal. That gap is the model’s own learned association.

These are statistical associations in the training data, not the model thinking.