INDEX

Explanations

Dutch and Netherlands

The neuron fires on country names and nationality terms (e.g. “Dutch,” “Netherlands,” “IJN”).

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

-2.02

<bos>

-1.63

of

-1.60

mış

-1.55

都

-1.47

</b>

-1.47

 первая

-1.47

↵↵

-1.46

 которая

-1.45

Pourquoi

-1.45

POSITIVE LOGITS

癭

1.69

 institu

1.63

ਃ

1.62

==>

1.59

荷兰

1.53

 enamor

1.52

埽

1.52

 fantasia

1.52

 assass

1.50

屺

1.50

Activations Density 0.171%