INDEX

Explanations

influence and change outcomes

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

$*$

-1.35

 erworben

-1.29

xv

-1.29

ſtion

-1.27

 دوس

-1.27

 ziren

-1.27

 aufgeführt

-1.26

 défis

-1.25

»),

-1.24

tj

-1.23

POSITIVE LOGITS

by

1.70

if

1.70

 eine

1.68

 even

1.62

of

1.55

in

1.55

for

1.49

one

1.41

 bringen

1.38

 виправи

1.36

Activations Density 0.033%