INDEX

Explanations

to

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

PyTuple

-0.48

υχ

-0.48

 bernama

-0.46

 forKey

-0.45

BagLayout

-0.43

versch

-0.42

 versch

-0.41

存于互联网档案馆

-0.41

 väli

-0.41

 przeci

-0.40

POSITIVE LOGITS

ⓧ

0.76

 achieving

0.74

 attaining

0.70

 winning

0.69

 onCancelled

0.69

the

0.67

 defeating

0.66

dubbo

0.66

 donating

0.66

 eliminating

0.66

Activations Density 0.130%