INDEX

Explanations

detection of threats and protection

describing or naming things

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 immerse

0.53

チョ

0.49

קור

0.48

トル

0.46

ysin

0.45

 immersed

0.44

noindent

0.44

꾿

0.43

nou

0.43

コロ

0.42

POSITIVE LOGITS

ib

0.52

Developing

0.50

廠商

0.49

ER

0.47

傷害

0.46

 maravilh

0.46

)*

0.46

靚

0.45

and

0.45

all

0.44

Activations Density 0.000%