INDEX

Explanations

increase or decrease

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

 kaarangay

-1.06

 kasarigan

-0.85

oa̍t

-0.81

 noqa

-0.76

parsedMessage

-0.76

ArrowToggle

-0.75

 للمعارف

-0.74

 дописавши

-0.73

CppCodeGen

-0.73

 autorytatywna

-0.72

POSITIVE LOGITS

 higher

1.24

higher

1.09

 lower

0.95

 HIGHER

0.91

Higher

0.91

 Higher

0.87

lower

0.73

 höher

0.72

 LOWER

0.64

 upper

0.63

Activations Density 0.001%