INDEX

Explanations

like or kind of

The neuron is keyed to detect standalone numeric tokens—especially multi‐digit and decimal numbers—wherever they appear.

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

 those

-1.47

it

-1.47

 delicado

-1.35

 colorida

-1.23

 đó

-1.16

}}$,

-1.15

resetPassword

-1.13

 encantador

-1.13

What

-1.13

 Ausblick

-1.13

POSITIVE LOGITS

for

1.31

já

1.20

while

1.18

nativeElement

1.15

misen

1.13

EDEN

1.11

 intensi

1.10

⎪

1.09

 traum

1.09

 constitu

1.08

Activations Density 0.052%