INDEX

Explanations

clean and dirty states

cleanliness and cleaning

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 pozycji

0.52

voz

0.45

alic

0.45

wit

0.42

ove

0.40

স্ফ

0.40

 চাল

0.40

 volat

0.40

ez

0.39

 помощ

0.39

POSITIVE LOGITS

 cleanliness

1.50

 unclean

1.47

 स्वच्छता

1.44

 dirty

1.43

dirty

1.38

 filthy

1.37

汚

1.34

 CLEAN

1.33

Clean

1.31

 clean

1.30

Activations Density 0.160%