INDEX

Explanations

light followed by flow

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

=",

1.68

 indifer

1.65

%",

1.58

 pikir

1.57

道府県

1.56

ophe

1.55

=?,

1.53

"=

1.53

hank

1.52

=",

1.52

POSITIVE LOGITS

ʈ

0.79

 pupils

0.75

 fauna

0.74

 composition

0.73

棕

0.73

 botany

0.73

 slate

0.72

幼儿

0.71

 students

0.71

writing

0.71

Activations Density 0.027%