INDEX

Explanations

presence of an idea

questions asked in a database

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ข

0.42

ﺘ

0.40

нг

0.39

 および

0.38

 schließlich

0.37

AKA

0.37

 Zusatz

0.37

事

0.36

כים

0.36

 Transmitter

0.36

POSITIVE LOGITS

 anymore

0.67

 than

0.66

into

0.65

’?

0.62

?”,

0.61

?”

0.60

 hanem

0.60

 into

0.59

?',

0.59

 onto

0.57

Activations Density 0.799%