INDEX

Explanations

the

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DockStyle

-0.96

DoubleQuotes

-0.79

setVerticalGroup

-0.78

Jeografia

-0.72

(!__

-0.67

GIVEREF

-0.67

 المعيارى

-0.66

TagMode

-0.66

Collegamenti

-0.65

AsUp

-0.65

POSITIVE LOGITS

 closely

0.51

 strongly

0.50

 deeply

0.46

/[

0.45

скар

0.45

In

0.45

 sincerely

0.43

Se

0.41

0.40

 centers

0.39

Activations Density 0.002%