INDEX

Explanations

URLs to software licenses

The neuron activates on URL components (e.g. “http://”, domain names, slashes, colons, dots), effectively detecting web addresses.

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

by

-1.46

not

-1.38

 just

-1.29

has

-1.27

 almost

-1.20

all

-1.17

 more

-1.16

 overtly

-1.14

 even

-1.14

one

-1.13

POSITIVE LOGITS

 เมื่อ

1.45

 quando

1.37

 การ์ตูน

1.29

beim

1.28

 também

1.28



1.27

 kabát

1.27

ientôt

1.27

ⓒ

1.27

腭

1.26

Activations Density 0.085%