INDEX

Explanations

titles and names

The neuron fires on personal honorifics (titles like Mr., Mrs., Ms., 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

戗

-1.36

够了

-1.35

🌰

-1.31

 kvalitet

-1.30

bakter

-1.30

ρος

-1.29

甗

-1.28

 sinyal

-1.28

独特的

-1.27

 любого

-1.27

POSITIVE LOGITS

 each

2.14

</b>

1.53

and

1.50

 because

1.36

 reduce

1.35

 what

1.34

 every

1.32

it

1.30

Activations Density 0.008%