INDEX

Explanations

references to the Kremlin or the Russian leadership

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

GIVEREF

-0.76

utilisons

-0.72

 AssemblyProduct

-0.69

Anhalt

-0.67

 Valerio

-0.67

tvguidetime

-0.66

 Jahn

-0.65

 архивлан

-0.65

 Vara

-0.65

المكان

-0.63

POSITIVE LOGITS

Kre

3.53

Kre

3.36

kre

2.41

kre

2.28

KRE

1.95

 Krebs

1.17

 Kremlin

1.09

 Кре

0.77

krebs

0.70

 Крем

0.62

Activations Density 0.001%