INDEX

Explanations

"the" capitalized word

This neuron fires on the word “the” when it introduces a formal name or title (i.e. the definite article before organization or entity names).

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

 with

-1.81

 similar

-1.78

Život

-1.63

 daarvoor

-1.50

 estrateg

-1.45

 those

-1.45

out

-1.39

lámpara

-1.38

 świecie

-1.38

 literally

-1.38

POSITIVE LOGITS

1.79

 theses

1.68

۰

1.63

ми

1.61

1.60

 dieser

1.59

↵↵

1.56

Spolja

1.55

};

1.53

1.51

Activations Density 0.206%