INDEX

Explanations

a or an followed by a word

The neuron is responsive to occurrences of the indefinite article “a” (and its close variants like “an,” “another”) and related small determiners.

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

 Then

-1.65

ījum

-1.63

 indicando

-1.56

"..."

-1.45

￭

-1.37

mêmes

-1.35

 كافة

-1.33

蔼

-1.33

種類の

-1.32

 lentamente

-1.31

POSITIVE LOGITS

欸

1.73

ra

1.66

Until

1.59

if

1.56

acquisto

1.54

ي

1.53

1.50

我很

1.50

的我

1.47

 racconta

1.45

Activations Density 0.034%