INDEX

Explanations

state indicators

This neuron detects CSS state‐class names (like “active,” “current,” “open,” etc.) in markup or style code.

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

of

-2.08

-1.91

-1.88

ке

-1.63

КЕ

-1.49

 either

-1.48

-1.45

下さ

-1.44

</i>

-1.40

 respectivas

-1.38

POSITIVE LOGITS

ﮑ

1.87

Probably

1.56

вгений

1.52

Semua

1.48

Fakta

1.46

توضیحات

1.46

Специ

1.44

Groet

1.43

Merhaba

1.41

Ещё

1.41

Activations Density 0.003%