INDEX

Explanations

questions beginning with what

what questions seeking details

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

 αφού

0.34

 ماشینونو

0.33

 сложных

0.32

ขึ้น

0.32

 όλα

0.32

 девя

0.31

 поез

0.31

 نیچے

0.31

 đốt

0.31

の時間

0.31

POSITIVE LOGITS

is

0.55

are

0.51

the

0.46

Are

0.45

 happened

0.42

re

0.41

}";

0.41

sthe

0.41

ta

0.40

Activations Density 0.165%