INDEX

Explanations

definition of a term

The neuron fires on uses of “called” when introducing a name or definition.

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

medriver

-1.65

aş

-1.59

ariate

-1.59

ajuku

-1.58

踌

-1.57

跻

-1.52

órios

-1.47

そこ

-1.47

Ꮸ

-1.46

titial

-1.44

POSITIVE LOGITS

by

2.16

1.66

“

1.64

슷

1.63

 berusaha

1.48

 dikel

1.43

 tersenyum

1.41

 mengangg

1.40

 綺麗

1.39

an

1.38

Activations Density 0.022%