INDEX

Explanations

image scaling and aspect ratio

The neuron activates on numeric values and terms that specify dimensions or measurements (e.g. width, height, aspect ratio, maxWidth, size, resizing).

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

 Himalaya

-0.94

 rất

-0.82

Où

-0.81

ϭ

-0.81

profesor

-0.80

 journalistes

-0.79

acherous

-0.78

municipi

-0.77

㎢

-0.77

Descriere

-0.77

POSITIVE LOGITS

 preserve

1.05

Preserve

1.03

new

1.02

new

1.00

 preserving

0.98

 target

0.92

 maintaining

0.92

 aspect

0.91

 keeping

0.91

preserving

0.90

Activations Density 0.014%