INDEX

Explanations

even numbers and vote

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

imshow

1.90

ized

1.86

্মী

1.82

ත්වය

1.81

 οποία

1.80

mise

1.77

ीण

1.76

 atteinte

1.72

 الجنوبية

1.72

 επιχει

1.68

POSITIVE LOGITS

ссо

1.61

 revolution

1.60

ろ

1.59

 бро

1.54

روع

1.53

axb

1.50

 heavy

1.49

 fresh

1.48

ров

1.48

ва

1.47

Activations Density 0.022%