INDEX

Explanations

counting

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

 chrétiens

-0.71

 démocr

-0.71

 généraux

-0.68

 normaux

-0.67

 fumée

-0.66

 religieuses

-0.66

 дописавши

-0.65

 venons

-0.65

 chré

-0.64

 titolata

-0.64

POSITIVE LOGITS

 count

0.67

 envol

0.65

 counts

0.62

 amount

0.60

sum

0.59

 correspond

0.57

 concern

0.56

AppMethodBeat

0.51

count

0.51

ContentLoaded

0.51

Activations Density 0.000%