INDEX

Explanations

camera finder / preview

The neuron selectively activates on occurrences of “finder” (especially in compounds like “viewfinder”) and related camera/optical finder terms.

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

قایناقلار

-0.79

ーブ

-0.75

welling

-0.73

Investigation

-0.73

 Clause

-0.72

leich

-0.70

ghal

-0.70

Serialization

-0.70

 walmart

-0.70

 Lancet

-0.69

POSITIVE LOGITS

finder

1.55

 finder

1.50

 Finder

1.34

 preview

1.32

Finder

1.17

 find

1.11

find

1.07

preview

1.03

FIND

1.03

 PREVIEW

1.01

Activations Density 0.034%