INDEX

Explanations

cities and countries

The neuron activates on bibliographic metadata—specifically publisher names and locations in reference or citation entries.

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

if

-1.31

Cartney

-1.14

～～

-1.02

 normas

-1.00

rý

-1.00

多摩

-0.97

=='

-0.95

 existi

-0.94

vaient

-0.93

 using

-0.92

POSITIVE LOGITS

 London

1.90

 Washington

1.72

 Tokyo

1.55

 Germany

1.53

 Italy

1.47

 Switzerland

1.46

 Paris

1.43

 Chicago

1.39

 Toronto

1.39

 Greece

1.36

Activations Density 0.007%