INDEX

Explanations

the

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

the

-0.55

Chham

-0.48

isEnd

-0.46

an

-0.45

 signaling

-0.44

stackSize

-0.42

łko

-0.41

ança

-0.40

ins

-0.40

POSITIVE LOGITS

 JpaRepository

0.77

 يتيمه

0.73

 متعلقه

0.68

attenzione

0.64

Tembelea

0.63

 hearts

0.63

TagMode

0.62

 consciences

0.61

 שוליים

0.60

Rhestr

0.59

Activations Density 0.002%