INDEX

Explanations

start of distinct phrases

The neuron is selectively detecting numeric literal tokens—especially floating‐point coordinates in an SVG path’s “d” attribute.

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

 partenariat

-1.41

 намного

-1.40

튿

-1.35

inizi

-1.31

Dislikes

-1.24

 améliorer

-1.24

жника

-1.22

 básicamente

-1.21

 gyer

-1.20

 labdar

-1.20

POSITIVE LOGITS

1.63

 their

1.51

1.48

for

1.42

now

1.41

 some

1.34

 ended

1.33

现在

1.31

was

1.30

.”

1.29

Activations Density 0.000%