INDEX

Explanations

privacy policies and consent

The neuron fires strongly on terms and phrases around protecting or keeping private sensitive information (especially medical/privacy contexts).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

to

-1.74

 with

-1.32

 arī

-1.23

but

-1.23

応え

-1.22

geordnet

-1.20

pão

-1.20

 because

-1.19

 estructura

-1.19

##

-1.16

POSITIVE LOGITS

 your

1.73

warnai

1.39

וּ

1.35

嘭

1.33

 السياس

1.28

ほら

1.27

ₒ

1.26

あぁ

1.24

嘻嘻

1.23

tifiable

1.23

Activations Density 0.127%