INDEX

Explanations

granting permission or access

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

 mereka

-1.82

Миха

-1.40

なのは

-1.39

也會

-1.36

等の

-1.34

by

-1.33

 Với

-1.33

 къде

-1.30

 Это

-1.30

൅

-1.30

POSITIVE LOGITS

 plenty

1.54

 into

1.45

 both

1.38

 atrav

1.31

㎏

1.25

 ample

1.24

änemark

1.23

ße

1.22

 zowel

1.21

 angusti

1.21

Activations Density 0.029%