INDEX

Explanations

Hart

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

 FetchType

-0.56

TagMode

-0.55

kloped

-0.50

TestBed

-0.50

$_['

-0.48

}{*}{}

-0.47

果た

-0.47

webElementXpaths

-0.47

WebElementEntity

-0.46

thest

-0.45

POSITIVE LOGITS

 dépens

0.61

 igång

0.60



0.59

nup

0.57

 avoient

0.56

 démocr

0.55

 parlent

0.54

 noires

0.54

BeginContext

0.53

 consommateurs

0.53

Activations Density 0.002%