INDEX

Explanations

qualities and characteristics

The neuron activates on descriptive modifiers—adjectives and adverbs (especially evaluative or degree‐marking words like “very,” “simple,” “low,” “important,” etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Шаг

-1.79

蛲

-1.73

摇了

-1.55

⌵

-1.52

 furent

-1.51

徬

-1.48

 studia

-1.48

 forse

-1.47

ávajú

-1.47

פורט

-1.45

POSITIVE LOGITS

 other

1.74

 would

1.73

it

1.65

but

1.60

 this

1.60

 will

1.52

 also

1.45

 information

1.42

1.41

 they

1.39

Activations Density 0.005%