INDEX

Explanations

China

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Error

-0.27

æĪªæŃ¢

-0.26

æ¦Ĩ

-0.26

éĹ²

-0.25

ä¼ł

-0.25

Angular

-0.24

bero

-0.24

az

-0.24

åľ°

-0.24

(Error

-0.24

POSITIVE LOGITS

meg

0.34

 swift

0.31

 foreground

0.31

meg

0.30

 massive

0.28

.CreateInstance

0.28

 giants

0.28

å¹´ä¸ŃåĽ½

0.27

dyn

0.26

å·¨äºº

0.26

Activations Density 0.887%