INDEX

Explanations

reputation

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

<bos>

-0.69

-0.57

らか

-0.54

 zodat

-0.54

 الأصل

-0.53

Билгалдахарш

-0.52

 whofe

-0.52

 Chriftian

-0.52

 Celui

-0.50

 leaſt

-0.50

POSITIVE LOGITS

AddTagHelper

0.67

ain

0.67

 DOES

0.66

LookAnd

0.66

is

0.63

 definitely

0.62

 equals

0.62

 DEFIN

0.61

can

0.60

isn

0.59

Activations Density 0.007%