INDEX

Explanations

doing or changing state

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

äº«

-0.30

 readily

-0.24

 easily

-0.23

ç«Łæĺ¯

-0.23

 appointments

-0.23

 Ð±Ð»Ð°Ð³Ð¾Ð´Ð°ÑĢÑı

-0.23

 bulund

-0.23

 thanks

-0.23

äº«åıĹ

-0.23

æīĵå¼ĢäºĨ

-0.22

POSITIVE LOGITS

æīįè¡Į

0.52

æīįèĥ½

0.38

æīįèĥ½å¤Ł

0.35

èĩ³å°ĳ

0.33

æīįåı¯ä»¥

0.33

åĲ¦åĪĻ

0.32

 somehow

0.31

æĸ¹åı¯

0.30

æīįæľī

0.30

è¿Ļæł·æīįèĥ½

0.29

Activations Density 0.035%