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

eclipse

-0.55

amazonaws

-0.50

keras

-0.50

msgTypes

-0.48

 gốc

-0.46

 רח

-0.45

ري

-0.44

ostro

-0.44

屈

-0.44

殊

-0.43

POSITIVE LOGITS

 الحره

0.63

)"),

0.62

Искәрмәләр

0.62

)];

0.60

 useAppContext

0.60

."],

0.60

 InputDecoration

0.59

 understanding

0.57

understanding

0.57

Studying

0.56

Activations Density 0.001%