INDEX

Explanations

Action words

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

æµģè¡ĮçļĦ

-0.27

é½¿è½®

-0.26

çļĦç©ºéĹ´

-0.26

loat

-0.25

æ¨¡

-0.25

 Äĳiá»ĥn

-0.25

åŀ¢

-0.24

modal

-0.24

çĶŁæ´»

-0.24

Wow

-0.24

POSITIVE LOGITS

 wrapping

0.33

ç»ķ

0.29

wrapper

0.29

 wrapper

0.28

èµ¶æĿ¥

0.27

 sling

0.27

åĬĿ

0.27

 jump

0.27

 bypass

0.26

è·³

0.26

Activations Density 0.891%