INDEX

Explanations

the word "combat"

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

-0.81

 TType

-0.71

 برانيه

-0.70

Kuz

-0.68

zio

-0.67

zulegen

-0.67

pexpr

-0.65

 matter

-0.65

RZ

-0.65

Sche

-0.63

POSITIVE LOGITS

 Combat

1.38

 combat

1.30

combat

1.30

Combat

1.26

 combats

0.95

 Kaufman

0.86

 combating

0.85

Personendaten

0.83

 kombat

0.81

 combatants

0.79

Activations Density 0.003%