INDEX

Explanations

downward direction and locations

downward direction or low position

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

 hoger

0.54

 höheren

0.51

 ऊंची

0.48

тория

0.47

 उच्च

0.47

の上

0.46

 আকাশে

0.46

 하늘

0.46

胀

0.46

 boven

0.45

POSITIVE LOGITS

bottom

1.27

 bottom

1.22

Bottom

1.18

 down

1.17

 پایین

1.17

 lower

1.15

 downwards

1.15

 below

1.14

 नीचे

1.13

 вниз

1.13

Activations Density 0.138%