INDEX

Explanations

local places and people

The neuron flags the main noun in “local X” references—i.e. it fires on the entity (students, department, family, hospital, company, etc.) that follows “local.”

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

its

-2.41

in

-1.65

 this

-1.65

 inside

-1.59

 than

-1.59

 from

-1.53

 jesús

-1.37

 itself

-1.31

poverty

-1.30

 quantità

-1.28

POSITIVE LOGITS

1.58

 will

1.53

 deinen

1.46

1.42

没有

1.38

the

1.34

da

1.30

your

1.30

Activations Density 0.040%