INDEX

Explanations

companionship and partnership

The neuron responds to words that denote a companion or partner (e.g., “companion,” “buddy,” “partner”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

-2.50

”？

-1.84

ayage

-1.77

 patriarchal

-1.57

</em>

-1.56

kaç

-1.55

 traditionelle

-1.49

 waarschijnlijk

-1.49

çok

-1.41

 …)

-1.40

POSITIVE LOGITS

在你

1.66

 miljoen

1.66

喜んで

1.63

1.59

卺

1.56

 with

1.55

derm

1.54

极了

1.53

ome

1.51

在我们

1.51

Activations Density 0.041%