INDEX

Explanations

full length

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 contempor

-0.85

Xs

-0.79

openConnection

-0.79

償

-0.78

 Auge

-0.75

peo

-0.72

gande

-0.71

ritten

-0.69

komen

-0.69

<bos>

-0.68

POSITIVE LOGITS

 full

1.06

 length

1.05

 immersion

0.90

 Full

0.87

Full

0.86

 FULL

0.86

 disclosure

0.85

①

0.84

 Länge

0.84

 Körper

0.83

Activations Density 0.022%