INDEX

Explanations

the abbreviation "Nat" possibly as part of a proper noun

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

iedział

-0.76

 Inno

-0.76

Prze

-0.75

NameInMap

-0.74

 Charlemagne

-0.72

 Kenne

-0.69

 Yeti

-0.69

Inno

-0.68

inno

-0.67

 Nossa

-0.67

POSITIVE LOGITS

Nat

1.68

Nat

1.48

nat

1.34

NAT

1.30

nat

1.25

NAT

1.15

 Natasha

0.92

Natasha

0.92

nats

0.91

abestanden

0.77

Activations Density 0.004%