INDEX

Explanations

even or perhaps qualifiers

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

of

-1.84

the

-1.53

him

-1.52

opies

-1.50

to

-1.50

but

-1.38

 แต่

-1.38

Its

-1.35

bo

-1.34

It

-1.31

POSITIVE LOGITS

philo

1.38

אל

1.32

parfum

1.30

ayangkan

1.29

argentina

1.27

rolex

1.27

$\$$

1.27

dcmtk

1.24

$*$

1.24

Latex

1.23

Activations Density 0.082%