INDEX

Explanations

email addresses

the presence of email addresses or web domains (hostname and top-level domain tokens).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 siècles

-1.51

 that

-1.38

␤

-1.36

вающие

-1.35

んだり

-1.33

вающий

-1.29

-1.28

='

-1.27

 derniers

-1.23

ächtige

-1.23

POSITIVE LOGITS

 would

1.55

Tarifs

1.31

 volont

1.30

The

1.24

ادت

1.20

 tentu

1.20

 TÉCNICA

1.18

できるので

1.18

’，

1.17

豋

1.16

Activations Density 0.065%