INDEX

Explanations

populated urban areas

The neuron detects language referring to high density or crowdedness—phrases that describe packed or heavily populated/trafficked environments.

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

</h3>

-1.38

 protože

-1.37

 novos

-1.36

ön

-1.36

 navegador

-1.34

prechung

-1.33

erráneo

-1.32

MLElement

-1.31

 Septembre

-1.31

`;

-1.30

POSITIVE LOGITS

it

1.44

 этих

1.43

 agregó

1.34

 kuitenkin

1.34

 Когда

1.34

 таких

1.33

 Эти

1.31

这是一

1.30

 inilah

1.28

がこの

1.27

Activations Density 0.035%