INDEX

Explanations

building shapes and descriptions

This neuron responds to technical architectural descriptors—particularly shape words (like “rectangular”), support/detail terms (“projecting,” “supported by”), and building measurements or numeric specifications.

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

 yrity

-2.02

 artikke

-1.99

Läh

-1.95

 tää

-1.90

 vektör

-1.87

ૉ

-1.84

蛱

-1.79

 سينما

-1.76

 viewHolder

-1.76

 ympä

-1.76

POSITIVE LOGITS

 パソコン

1.63

散文

1.55

 встроен

1.49

อยู่

1.48

is

1.45

 lojas

1.43

 logements

1.40

 receta

1.38

 oreja

1.38

Activations Density 0.152%