INDEX

Explanations

specific starting phrases

This neuron fires on banner‐style headings or promotional slogans (e.g. attention‐grabbing titles and calls to action).

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

 ALONE

-1.38

 smal

-1.32

 verminder

-1.26

вому

-1.26

 encontram

-1.24

 hemd

-1.22

fyr

-1.21

𞤢

-1.20

⇏

-1.19

 bringe

-1.17

POSITIVE LOGITS

to

2.14

</h3>

1.59

be

1.40

 still

1.39

";

1.39

so

1.38

now

1.38

 other

1.30

("

1.30

աստ

1.30

Activations Density 0.126%