INDEX

Explanations

place names and administrative divisions

The neuron primarily activates on commas (and similar boundary punctuation) that separate names, places, or list items.

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

 from

-1.84

 this

-1.52

astră

-1.47

 different

-1.41

celotti

-1.37

⸒

-1.34

bės

-1.32

 beda

-1.31

 rossi

-1.30

νης

-1.30

POSITIVE LOGITS

1.63

 accentuated

1.55

↵

1.54

 каждого

1.41

of

1.39

 слишком

1.39

 estekak

1.31

릭터

1.30

 другие

1.29

 arriba

1.27

Activations Density 0.178%