INDEX

Explanations

universal words

The neuron fires on sentence-initial discourse connectors or transition words (often inside quotes), e.g. “It,” “also,” “Thus,” “Combined with,” etc.

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

 being

-1.78

埛

-1.61

it

-1.52

 überprü

-1.47

-------

-1.46

ूरत

-1.43

 such

-1.40

 said

-1.38

艽

-1.38

can

-1.37

POSITIVE LOGITS

all

1.88

 все

1.64

全て

1.59

all

1.56

for

1.52

 všetky

1.48

 semua

1.46

 جميع

1.46

 leurs

1.45

 всі

1.45

Activations Density 0.184%