INDEX

Explanations

languages and specific terms

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

 helfen

1.42

 keinen

1.39

 effektiv

1.36

 definitiv

1.31

 weltweit

1.30

 optimal

1.29

 completely

1.27

 inherent

1.25

 yearly

1.25

 generell

1.25

POSITIVE LOGITS

 শ্যাম

1.66

Nathan

1.66

 повідом

1.64

记者

1.62

 الدكتور

1.60

 นาย

1.53

わたし

1.50

述べ

1.48

 สาว

1.47

 Lucas

1.46

Activations Density 0.054%