INDEX

Explanations

restored, renovated, refurbished

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

any

-1.70

 которая

-1.59

 expresó

-1.51

是吗

-1.50

 quienes

-1.46

 utilizan

-1.45

 Nhưng

-1.45

錚

-1.43

明確

-1.41

 Einige

-1.38

POSITIVE LOGITS

 daga

1.72

—”

1.70

会

1.66

.";

1.58

ALREADY

1.54

 complètement

1.53

 ſont

1.52

↵

1.51

 leone

1.51

‘

1.50

Activations Density 0.035%