INDEX

Explanations

UI, panels, canvas

The neuron flags code‐style tokens—especially programming keywords and identifiers (e.g. public, void, override, GUI, InputField, Unity)—i.e. it detects code snippets.

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

当選

-0.80

 Tagged

-0.79

ListFragment

-0.78

Isa

-0.75

Ton

-0.75

 palmar

-0.75

świad

-0.73

userAgent

-0.72

SuppressWarnings

-0.71

СС

-0.70

POSITIVE LOGITS

UI

1.45

 panel

1.41

Canvas

1.35

 Canvas

1.23

Panel

1.21

 panels

1.21

 canvas

1.21

UI

1.20

 image

1.20

ui

1.18

Activations Density 0.003%