INDEX

Explanations

which

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

delig

-0.47

 prove

-0.46

 biztos

-0.45

 như

-0.44

知

-0.44

-0.43

ответ

-0.42

 help

-0.42

are

-0.41

 φα

-0.41

POSITIVE LOGITS

TagMode

0.98

########.

0.96

 disambiguazione

0.92

 Monfieur

0.83

elemField

0.79

Efq

0.78

 Majefty

0.78

styleType

0.77

ScopeManager

0.73

 GenerationType

0.70

Activations Density 0.001%