INDEX

Explanations

little prefixes

The neuron fires on roman‐letter chunks of Chinese names or terms (i.e. Pinyin syllables).

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

 Conſ

-1.61

 motocicletas

-1.56

 údaje

-1.55

 remboursement

-1.54

㌧

-1.53

 listos

-1.52

 cobr

-1.44

disfraz

-1.41

 produjo

-1.40

߹

-1.39

POSITIVE LOGITS

sistence

1.63

do

1.59

 }));

1.53

':'

1.52

</i>

1.52

葙

1.51

桫

1.51

ᨅ

1.50

merate

1.50

uldade

1.50

Activations Density 0.006%