INDEX

Explanations

court case names

The neuron activates on numeric tokens, i.e. digit sequences or numbers in the text.

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

or

-1.68

 سينما

-1.57

to

-1.57

 motocicletas

-1.56

 प्रोड

-1.48

 kontaktu

-1.48

 reivindic

-1.48

صبحت

-1.43

 cukru

-1.42

 土産

-1.39

POSITIVE LOGITS

 then

1.68

 السياس

1.67

ente

1.52

ּוֹ

1.48

dig

1.47

 présents

1.47

ַּ

1.46

ђа

1.44

1.42

之事

1.41

Activations Density 0.026%