INDEX

Explanations

capitalized words, names, and locations, especially in the context of news events

locations

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

Negative Logits

swt

-0.52

 lepas

-0.48

روسية

-0.47

 sẻ

-0.46

ActivityResult

-0.46

 lateral

-0.45

METADATA

-0.45

 نيوز

-0.44

orgs

-0.44

 Pugh

-0.43

POSITIVE LOGITS

tagHelperRunner

0.86

 مرئيه

0.76

########.

0.75

 Wikimedijinoj

0.69

OGND

0.68

AndEndTag

0.68

yntaxException

0.66

matchCondition

0.65

المكان

0.65

PerformLayout

0.62

Activations Density 0.246%