INDEX

Explanations

code block closing brace }

The neuron is detecting numeric literals (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

 Nieuws

-1.51

暃

-1.46

 massiv

-1.45

__

-1.45

 faciliter

-1.42

ᐉ

-1.41

毉

-1.37

 Dingen

-1.36

 puits

-1.35

éric

-1.33

POSITIVE LOGITS

 using

1.73

so

1.66

for

1.62

has

1.59

or

1.51

 because

1.51

1.38

 based

1.37

in

1.34

1.33

Activations Density 0.009%