INDEX

Explanations

infiltration

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 EnglishChoose

-0.71

posedge

-0.71

Hentet

-0.69

 Physically

-0.68

Personendaten

-0.67

 infiltrate

-0.65

IndentedString

-0.65

ValueStyle

-0.65

 infiltration

-0.65

 physically

-0.64

POSITIVE LOGITS

0.77

 együtt

0.60

 Juifs

0.56

 investissements

0.56

 egne

0.55

 čier

0.54

 törvény

0.54

værende

0.53

 čierna

0.53

tli

0.53

Activations Density 0.091%