INDEX

Explanations

truth or current state

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

of

-2.30

 then

-1.88

for

-1.75

in

-1.68

on

-1.62

外的

-1.42

by

-1.38

if

-1.35

-1.34

 more

-1.34

POSITIVE LOGITS

]]);

1.55

GOTREF

1.52

 Walkover

1.49

鲧

1.48

傒

1.45

compromised

1.41

 Seeder

1.39

齶

1.39

hlte

1.39

 Adven

1.38

Activations Density 0.087%