INDEX

Explanations

scandal

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

picious

-0.81

 חיצוניים

-0.72

ologists

-0.69

ftagPool

-0.67

tly

-0.65

ainted

-0.63

yards

-0.60

RenderAtEndOf

-0.59

tableFuture

-0.58

errHandler

-0.58

POSITIVE LOGITS

 lèvres

0.43

 préférences

0.42

 nature

0.40

 religione

0.40

__(

0.40

間に

0.40

 çöz

0.40

Læs

0.39

vég

0.38

 claim

0.38

Activations Density 0.212%