INDEX

Explanations

construction projects and geometry

The main thing this neuron does is find mentions of construction or building activities.

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

 muzzle

0.78

(.*

0.78

 Packet

0.75

 Puck

0.73

 Glycer

0.73

kDa

0.73

Πα

0.73

Provider

0.72

 berries

0.71

Bearer

0.71

POSITIVE LOGITS

建築

1.53

建筑

1.48

 건축

1.48

 Skyscrapers

1.45

🏗

1.42

buildings

1.41

 architectural

1.37

建設

1.35

 건설

1.34

 buildings

1.33

Activations Density 1.676%