INDEX

Explanations

file paths and dates

The neuron fires on numeric tokens (digits, slashes, and dots) that form date or version‐style strings.

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

-1.68

 zdarma

-1.61

ﹶ

-1.60

Actually

-1.54

를

-1.43

𓆉

-1.39

 apparaître

-1.39

-1.38

Definitely

-1.34

POSITIVE LOGITS

 first

1.49

ıdır

1.41

of

1.36

it

1.31

 students

1.30

 büyük

1.27

 svoju

1.26

 which

1.25

 most

1.25

 every

1.23

Activations Density 0.021%