INDEX

Explanations

what is questions

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

åģļä¸ĢäºĽ

-0.29

uria

-0.28

ogo

-0.28

 rÃ©alitÃ©

-0.27

done

-0.27

 done

-0.25

çİ°å®ŀä¸Ń

-0.24

åİ»åģļ

-0.24

 Done

-0.24

possible

-0.24

POSITIVE LOGITS

åı¦ä¸Ģä¸ª

0.47

another

0.40

 another

0.39

Another

0.38

one

0.37

 Ã©Ã©n

0.37

One

0.37

-One

0.34

ä¹ĭä¸Ģ

0.34

åı¦ä¸Ģ

0.34

Activations Density 0.010%