INDEX

Explanations

% followed by a word

The neuron is specialized to detect code tokens that immediately follow a percent sign (i.e. “%” magic or directive keywords).

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

喾

-1.63

ዡ

-1.60

鼕

-1.53

 esteemed

-1.50

pendidikan

-1.48

 detaylı

-1.41

EREN

-1.39

 adecuados

-1.38

 famed

-1.38

and

-1.38

POSITIVE LOGITS

 grün

1.57

’,

1.46

й

1.38



1.37

 skolan

1.37

Donde

1.35

 Edelstahl

1.35

Cómo

1.35

Otras

1.34

sandalia

1.33

Activations Density 0.023%