INDEX

Explanations

explaining to me why

This neuron activates on phrases where someone is communicating or giving information—i.e. verbs like “share,” “explain,” or “instill” paired with a following preposition (“with,” “to,” “in”) and its object.

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

-2.16

-1.77

 häufig

-1.69

 Когда

-1.63

 respaldo

-1.63

 neuen

-1.55

-1.54

 его

-1.52

because

-1.48

POSITIVE LOGITS

ússia

1.84

戋

1.84

 geene

1.80

۶

1.79

and

1.78

肫

1.77

۷

1.73

tembre

1.62

۹

1.61

 ۱۵

1.59

Activations Density 0.024%