INDEX

Explanations

conjunctions and punctuation

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

aminer

-0.29

[["

-0.27

stood

-0.27

atern

-0.26

 dint

-0.25

verty

-0.24

 stocked

-0.24

çĹĶ

-0.24

ilt

-0.24

akens

-0.24

POSITIVE LOGITS

ÐĵÐ»Ð°Ð²Ð½Ð°Ñı

0.29

é¦ĸåıĳ

0.27

tet

0.26

å¤ĩæ³¨

0.26

ç¬¬ä¸Ģå¤©

0.24

xbe

0.24

HOME

0.24

å±ħå®¶

0.24

æĹĭ

0.23

çĶŁåĳ½çļĦ

0.23

Activations Density 0.002%