INDEX

Explanations

frequency

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

are

-0.63

don

-0.54

 happen

-0.54

end

-0.52

 aren

-0.51

 tend

-0.50

 always

-0.49

-0.48

all

-0.47

POSITIVE LOGITS

SharedDtor

0.84

astéroïdes

0.83

 poffe

0.82

berdayakan

0.82

 pleaſure

0.79

featureID

0.79

 itſelf

0.79

 Monfieur

0.78

 Jefus

0.77

 Majefty

0.77

Activations Density 0.075%