INDEX

Explanations

streets and districts

The neuron is looking for mentions of specific urban geographic locations or administrative areas—words like “district,” “street,” “avenue,” “boulevard,” “capital,” or “center.”

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

或

-1.52

หรือ

-1.47

 یا

-1.42

 because

-1.40

when

-1.39

hodné

-1.37

 innych

-1.33

或者

-1.32

 или

-1.32

They

-1.32

POSITIVE LOGITS

1.73

1.70

1.48

猖

1.44

1.43

 busiest

1.32

 hottest

1.28

 permitem

1.27

 permite

1.26

$\

1.25

Activations Density 0.044%