INDEX

Explanations

based in locations

The neuron detects mentions of an entity’s geographic base—i.e. phrases like “based in” or “based at” followed by a location.

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

別人

-1.89

虯

-1.64



-1.59

艫

-1.53

</td>

-1.52

 soigneusement

-1.51

鰱

-1.51

 Depuis

-1.47

 animali

-1.46

 большо

-1.45

POSITIVE LOGITS

at

1.71

 latter

1.65

in

1.59

out

1.52

‘

1.49

to

1.45

ll

1.38

artament

1.38

in

1.38

up

1.37

Activations Density 0.037%