INDEX

Explanations

want followed by you, other, only, or to

The neuron fires on occurrences of the word “want” (especially in “we want” constructions) indicating a stated desire or request.

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

we

-1.62

nossa

-1.46

umumkan

-1.42

前面

-1.41

 笑顔

-1.40

ıyordu

-1.40

 being

-1.39

for

-1.39

see

-1.38

salud

-1.38

POSITIVE LOGITS

 "}";

1.68

 THOSE

1.48

/>}

1.48

 "../../../../

1.45

_));

1.43

ᚔ

1.42

vello

1.40

DataMember

1.40

わざわざ

1.40

 ['./

1.38

Activations Density 0.031%