INDEX

Explanations

why do

explaining concepts and actions

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

 आजा

0.52

 arcu

0.47

 govor

0.46

GALAD

0.46

 olursa

0.46

 使っ

0.46

 više

0.45

вик

0.45

ம்பர்

0.45

 почина

0.45

POSITIVE LOGITS

We

0.57

 Diagram

0.54

Мы

0.52

 They

0.51

 Diagrams

0.49

0.48

 Broad

0.47

We

0.47

re

0.46

Activations Density 0.000%