INDEX

Explanations

purpose or goal

The neuron detects wording that introduces an entity’s objectives—terms like “purpose,” “mission,” “aim,” or “goal” that state what something is set up to do.

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

 nonatomic

-1.98

 them

-1.75

していない

-1.64

 things

-1.59

したもの

-1.55

してない

-1.53

 informacje

-1.52

our

-1.52

するもの

-1.51

好像是

-1.50

POSITIVE LOGITS

is

2.08

1.61

and

1.55

.”

1.51

1.40

1.31

 servent

1.31

##

1.29

dienne

1.28

1.26

Activations Density 0.086%