INDEX

Explanations

wp, files, node addresses

The neuron activates on URL‐style path components and file/directory names (e.g. “wp‐content,” “plugins,” “modules,” “node,” “taxonomy,” etc.) in text.

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

 from

-1.58

and

-1.48

"""

-1.47

 illustratie

-1.42

 while

-1.31

 همچ

-1.31

};

-1.28

});

-1.27

 relevancia

-1.26

 takie

-1.25

POSITIVE LOGITS

Karakteristik

1.60

 ferait

1.52

賔

1.44

Ventajas

1.44

Tratamiento

1.39

Beneficios

1.39

Siga

1.38

Reprodução

1.32

这对

1.32

的很

1.32

Activations Density 0.019%