INDEX

Explanations

for all practical purposes

for demonstration or clarity

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 fordi

0.25

uem

0.24

ι

0.24

이라

0.23

en

0.23

also

0.23

 Also

0.23

es

0.23

''

0.23

\'

0.23

POSITIVE LOGITS

 starters

0.57

erun

0.57

 purposes

0.57

 example

0.53

 instance

0.53

 reasons

0.52

 awhile

0.52

geries

0.48

giveness

0.48

erunner

0.48

Activations Density 0.836%