INDEX

Explanations

word followed by phrase

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

μος

1.59

KAN

1.50

 принци

1.50

unik

1.47

cored

1.43

μου

1.42

 кана

1.40

 теп

1.38

कोर

1.38

сса

1.38

POSITIVE LOGITS

 hydrated

1.63

 hourglass

1.54

 guiclass

1.53

 almac

1.50

 observable

1.48

 flush

1.46

 işlem

1.46

 transmembrane

1.43

 actionBar

1.40

 abundance

1.39

Activations Density 0.000%