INDEX

Explanations

cities and their context

The neuron detects and strongly activates on capitalized geographic place names (especially cities).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

บ

0.60

 rearrange

0.59

フィルタ

0.57

笊

0.56

λλη

0.55

同学们

0.55

চক

0.55

该

0.55

該

0.55

POSITIVE LOGITS

shire

1.32

 Üniversitesi

1.30

 City

1.29

 Selatan

1.27

광역시

1.27

 County

1.22

 suburb

1.22

 Airport

1.20

州的

1.17

州

1.16

Activations Density 1.069%