INDEX

Explanations

improved safety

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

();)

-0.62

stretchr

-0.60

 vectorielles

-0.57

RegressionTest

-0.56

 kuin

-0.54

 mesmas

-0.52

 soggetto

-0.50

 singuli

-0.49

 than

-0.47

 sujets

-0.47

POSITIVE LOGITS

AndEndTag

0.67

 externi

0.63

ity

0.60

ific

0.59

Билгалдахарш

0.58

ajaj

0.58

ities

0.57

dci

0.57

NameInMap

0.56

rific

0.56

Activations Density 0.005%