INDEX

Explanations

proprietary information

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

-2.25

ずつ

-1.56

by

-1.55

 vuotta

-1.50

 всех

-1.43

ﾈ

-1.41

넣

-1.39

тельном

-1.34

四个

-1.30

 hinsichtlich

-1.27

POSITIVE LOGITS

：

1.47

1.46

 dalamnya

1.39

Apologies

1.38

⇹

1.35

 lorsqu

1.33

 they

1.29

 représenter

1.28

 ultimi

1.27

yā

1.27

Activations Density 0.012%