INDEX

Explanations

first, then, followed by

The neuron fires on step‐indicator or sequence‐marker words (especially “First” and other tokens that open numbered or ordered steps in a procedure).

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

 also

-1.51

 також

-1.41

 גם

-1.37

 επίσης

-1.27

 также

-1.20

Dès

-1.17

もありました

-1.16

也会

-1.16

 particularly

-1.13

 myös

-1.11

POSITIVE LOGITS

 først

1.02

1.00

 ثم

0.98

 ensin

0.95

先

0.91

nių

0.89

 fresca

0.87

Then

0.87

 Followed

0.86

 then

0.86

Activations Density 0.044%