INDEX

Explanations

maps to diverse types

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

 möjlighet

-1.56

EdgeInsets

-1.49

uminense

-1.48

爱心

-1.47

凝土

-1.44

ih

-1.43

 můžete

-1.42

rália

-1.42

硖

-1.41

ęgn

-1.41

POSITIVE LOGITS

the

2.64

Зміст

1.74

 only

1.41

 έχουν

1.40

 für

1.39

 Thus

1.37

 매우

1.37

in

1.36

 Regarding

1.33

 what

1.30

Activations Density 0.010%