INDEX

Explanations

capture or rejection

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 يتيمه

-1.05

 itſelf

-0.83

 للمعارف

-0.80

OGND

-0.79

Efq

-0.77

rungsseite

-0.76

 purpoſe

-0.76

 houſe

-0.75

SequentialGroup

-0.75

Eſ

-0.75

POSITIVE LOGITS

in

0.58

 lisäksi

0.50

to

0.49

 crossorigin

0.48

0.46

0.43

 řízení

0.42

ora

0.41

...

0.41

GUILayout

0.41

Activations Density 0.075%