INDEX

Explanations

meaning 'woman' or 'land'

The neuron activates on words given as literal glosses or translations—i.e. the quoted or parenthetical “meaning” terms in etymology/definition sections.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

̞

-1.28

'"

-1.21

Müller

-1.17

pozycji

-1.17

 الاسلام

-1.13

isierung

-1.13

"-

-1.13

 כדי

-1.12

ógicas

-1.12

<0xCE>

-1.12

POSITIVE LOGITS

for

1.84

 when

1.49

as

1.41

 these

1.40

 most

1.33

of

1.31

ouvre

1.30

 melewati

1.30

 they

1.28

 also

1.23

Activations Density 0.046%