INDEX

Explanations

types of buildings

The neuron detects mentions of specific building or structure types (e.g., house, cottage, shed, garage, boathouse, pavilion, etc.).

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

 kwamba

-1.52

вгений

-1.47

 nyingi

-1.38

⭔

-1.37

 eenig

-1.34

鄚

-1.34

 puder

-1.33

妧

-1.32

 Polda

-1.30

うわ

-1.30

POSITIVE LOGITS

It

1.68

It

1.63

was

1.60

 There

1.41

 built

1.37

 used

1.32

1.30

Of

1.28

 With

1.27

By

1.23

Activations Density 0.044%