INDEX

Explanations

For

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

 незавершена

-0.49

ksikon

-0.45

ItemBackground

-0.45

RegressionTest

-0.44

伝わ

-0.41

writeFieldEnd

-0.41

may

-0.39

VertexAttrib

-0.39

त

-0.39

RuleContext

-0.39

POSITIVE LOGITS

 سكانية

0.68

 AssemblyCompany

0.66

 aught

0.66

 chrétienne

0.65

UnsafeEnabled

0.64

<!--[

0.62

cible

0.61

ตร์

0.60

Хочу

0.60

 كومونز

0.59

Activations Density 0.042%