INDEX

Explanations

photography services and subjects

The neuron is responding to numeric tokens—especially multi‐digit sequences like years or other date/metadata numbers.

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

as

-1.48

 long

-1.42

 sound

-1.34

 regarding

-1.30

 values

-1.29

 health

-1.27

 children

-1.25

 large

-1.22

 economic

-1.22

-1.21

POSITIVE LOGITS

嶮

1.41

龇

1.34

}%

1.33

 チラ

1.25

requently

1.25

枼

1.20

 ホルダー

1.20

$_=

1.20

ܲ

1.20

Ostat

1.19

Activations Density 0.260%