INDEX

Explanations

substantially reduce

multilingual validity checks

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 click

0.50

 Phillips

0.50

 pets

0.48

 renewable

0.48

 Crucible

0.48

 Tale

0.48

 creature

0.47

 Farmington

0.47

 Roofing

0.47

 psal

0.47

POSITIVE LOGITS

 destruir

0.50

ਬ

0.49

ينه

0.46

菪

0.46

わけ

0.46

 মাংস

0.46

 формы

0.45

賤

0.45

总部

0.43

يتها

0.43

Activations Density 0.001%