INDEX

Explanations

have to provide

affecting consequences and outcomes

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

0.55

 general

0.54

 future

0.54

 sword

0.53

 module

0.52

 tower

0.51

',

0.51

men

0.51

POSITIVE LOGITS

pOt

0.57

aat

0.54

Daten

0.52

aap

0.52

Smooth

0.50

oupe

0.49

ρικ

0.48

 representación

0.48

AppLink

0.47

 நடிப்ப

0.47

Activations Density 0.001%