INDEX

Explanations

force management

The neuron fires on occurrences of the root “force” (in any case or morphological form, e.g. force, Force, forced, FORCE).

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

 phare

-1.92

簠

-1.80

 frança

-1.80

:))

-1.78

 secr

-1.77

???????

-1.77

躂

-1.77

opportunity

-1.76

甞

-1.75

ጐ

-1.75

POSITIVE LOGITS

of

2.42

et

2.31

is

1.90

ripsi

1.88

1.85

 не

1.75

—

1.73

don

1.71

or

1.66

to

1.66

Activations Density 0.024%