INDEX

Explanations

hard

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

Ton

-0.30

æĢĢæĬ±

-0.29

æĺ¾èĳĹ

-0.28

eous

-0.28

Mot

-0.28

çªĹå¤ĸ

-0.27

ä¸įå°ıçļĦ

-0.27

 significant

-0.25

é²ľèī³

-0.25

ç»Ł

-0.25

POSITIVE LOGITS

aland

0.32

ibraries

0.27

ç¡¬

0.26

aldo

0.25

èµ·æĿ¥

0.24

iest

0.24

etta

0.24

å®ĮäºĨ

0.24

è¢Ń

0.24

ensa

0.24

Activations Density 0.884%