INDEX

Explanations

code structure delimiters

The neuron is detecting numeric tokens (numbers and numeric literals) 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

<bos>

-1.93

by

-1.90

in

-1.83

 because

-1.75

to

-1.64

 there

-1.54

welli

-1.34

は

-1.33

 Infatti

-1.33

razine

-1.32

POSITIVE LOGITS

に入れる

1.65

だいぶ

1.64

にします

1.63

無論

1.61

もうすぐ

1.59

あー

1.59

のでしょう

1.58

ほら

1.57

んでしょう

1.55

いきなり

1.55

Activations Density 0.000%