INDEX

Explanations

things happening

The neuron identifies the main subject or topic noun of each question (e.g. “site,” “app,” “file,” “board,” “game,” “stock,” “car,” “program”).

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

of

-2.70

anging

-1.45

 useAppContext

-1.40

。”

-1.37

-1.36

γραφία

-1.33

but

-1.28

-1.27

 camiset

-1.27

鹇

-1.26

POSITIVE LOGITS

 altså

1.39

 поддер

1.39

 marcada

1.30

 necesarios

1.28

 ifølge

1.27

The

1.27

了不少

1.27

 frág

1.23

 Teile

1.23

を作る

1.21

Activations Density 0.242%