INDEX

Explanations

type or style

The neuron is tuned to detect occurrences of the field label “Type” (especially in “Type of …”) in structured document headers or forms.

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

 therefore

-1.65

 have

-1.57

舠

-1.43

 needed

-1.36

 deshalb

-1.34

 hvilke

-1.34

~”

-1.33

-1.30

 hvorfor

-1.30

ﮩ

-1.28

POSITIVE LOGITS

1.59

力和

1.30

その他

1.30

心の

1.29

enzuela

1.28

 např

1.28

ioi

1.27

makaian

1.26

 original

1.24

1.23

Activations Density 0.035%