INDEX

Explanations

credit future right

"it" followed by punctuation or specific words

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

 murderous

0.61

्ठा

0.57

дной

0.54

 rampage

0.52

 constitut

0.51

튜브

0.51

 baterías

0.49

 foodservice

0.48

 quadruple

0.47

チューブ

0.46

POSITIVE LOGITS

 Screening

0.48

转

0.48

叁

0.47

 स्क्रीनिंग

0.46

 حالا

0.45

 ση

0.44

整

0.44

ene

0.43

屏幕

0.43

—

0.43

Activations Density 0.001%