INDEX

Explanations

whiskers or filament

The neuron primarily detects numeric literals and numeric expressions (e.g. version numbers, coordinates, dimensions, scale factors).

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

ּוֹ

-1.88

brities

-1.36

ִּ

-1.34

וּ

-1.33

 бампер

-1.25

וֹ

-1.16

нци

-1.14

ırım

-1.13

ַּ

-1.11

Unternehmen

-1.08

POSITIVE LOGITS

ׇ

1.56

 inicios

1.13

 camioneta

1.13

 αὐ

1.11

 incluso

1.08

1.07

 semej

1.03

=[];

1.02

 Einstellungen

1.02

 zwierząt

1.02

Activations Density 0.007%