INDEX

Explanations

suppose

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

 resourceCulture

-0.85

ised

-0.83

nologue

-0.81

Geplaatst

-0.75

InjectAttribute

-0.68

LayoutStyle

-0.67

shafen

-0.67

 gräns

-0.67

ThemeOverlay

-0.67

 createSlice

-0.66

POSITIVE LOGITS

up

0.45

0.41

ers

0.38

 altogether

0.38

 from

0.38

 personally

0.37

 back

0.37

at

0.36

est

0.36

ats

0.36

Activations Density 1.755%