INDEX

Explanations

if followed by potential outcome

The neuron activates on proper personal names (mentions of people).

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

 Organisateur

-1.31

stehen

-1.30

 rése

-1.30

 altijd

-1.30

 alltid

-1.28

 incrí

-1.28

 Pristup

-1.27

">“

-1.26

......”

-1.26

 igra

-1.24

POSITIVE LOGITS

any

2.20

get

1.69

 continue

1.65

not

1.63

 anything

1.48

as

1.46

go

1.42

 truly

1.39

1.38

 loses

1.38

Activations Density 0.014%