INDEX

Explanations

Before

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

oy

-0.23

è¾ĩ

-0.21

 campaigners

-0.21

 underrated

-0.20

 unmist

-0.20

 subpo

-0.20

memberof

-0.19

 nomine

-0.19

 imprison

-0.19

 resembl

-0.19

POSITIVE LOGITS

HAND

0.28

ruise

0.28

hand

0.25

 createStackNavigator

0.25

:".$

0.23

uisine

0.23

/post

0.23

eworld

0.22

chedulers

0.22

beeld

0.22

Activations Density 1.025%