INDEX

Explanations

the word "be"

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

сім

-0.68

います

-0.61

hezza

-0.60

 marrone

-0.59

 ignore

-0.57

clearTimeout

-0.56

 forgot

-0.54

看看

-0.54

polate

-0.54

 proposito

-0.54

POSITIVE LOGITS

 being

1.73

being

1.63

 BEING

1.59

Being

1.55

 Being

1.49

 having

1.09

 bein

1.06

having

1.05

être

1.03

 essendo

1.03

Activations Density 0.428%