INDEX

Explanations

Crosse

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

accurate

-1.03

 accurate

-0.93

Accurate

-0.91

urate

-0.82

adors

-0.80

leggen

-0.78

 Accurate

-0.76

 nhàng

-0.75

BibitemShut

-0.74

umptions

-0.66

POSITIVE LOGITS

 jurídica

0.47

Demografia

0.46

spection

0.46

 StatefulWidget

0.45

BeginContext

0.41

hr

0.41

 muualla

0.40

dL

0.40

VersionUID

0.40

 gestation

0.40

Activations Density 0.037%