INDEX

Explanations

concern

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

insee

-0.56

 مرئيه

-0.54

 Meksiku

-0.54

 Paglinawan

-0.51

WebElementEntity

-0.49

ecd

-0.48

 Catawiki

-0.47

lüğ

-0.47

蚪

-0.47

 netto

-0.45

POSITIVE LOGITS

SBATCH

0.66

 itſelf

0.63

perative

0.59

.";

0.57

 priser

0.56

 køb

0.56

 Chriſt

0.55

 Allegretto

0.55

 Reſ

0.55

 næste

0.54

Activations Density 0.017%