INDEX

Explanations

"the" followed by proper nouns or specific entities

The neuron consistently activates on the definite article “The.”

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

ztes

-1.80

время

-1.70

biendo

-1.66

絝

-1.63

),

-1.63

 keď

-1.61

プリング

-1.59

 když

-1.57

 referrerpolicy

-1.52

 píše

-1.52

POSITIVE LOGITS

of

2.48

 this

2.23

this

2.17

 ayakkabı

1.65

後は

1.64

 such

1.55

 pères

1.55

 trofeo

1.52

 innym

1.50

1.49

Activations Density 0.138%