INDEX

Explanations

through

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

 below

-0.49

 across

-0.49

no

-0.48

σχ

-0.47

in

-0.47

 through

-0.46

 moeite

-0.43

to

-0.43

 passando

-0.41

참고

-0.40

POSITIVE LOGITS

rawDesc

0.85

TagHelper

0.68

esModule

0.64

complexContent

0.64

 Ranks

0.63

jectable

0.61

argout

0.61

出版年

0.61

adə

0.60

principalColumn

0.60

Activations Density 0.012%