INDEX

Explanations

crowd of

The neuron detects mentions of groups of spectators or audiences (e.g., “crowd,” “audience,” “spectators”).

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

now

-1.87

 şimdi

-1.51

meteries

-1.46

 then

-1.45

 many

-1.35

Now

-1.35

 that

-1.32

 about

-1.27

自

-1.27

end

-1.26

POSITIVE LOGITS

Cuando

1.41

かけた

1.35

则是

1.34

 polaire

1.33

 möjlighet

1.33

뭘

1.30

賅

1.30

 című

1.28

ﮯ

1.28

jdzie

1.26

Activations Density 0.031%