INDEX

Explanations

UI elements, dialogs, and forms

The neuron fires on code identifiers—especially CamelCase names of classes or UI components (e.g. “SERVER”, “protocol”, “widget”, “Dlg”, “Window”, “Form”).

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

its

-1.23

+*

-1.16

on

-1.16

 miljoen

-1.11

...@

-1.09

or

-1.09

)：

-1.08

 have

-1.07

strator

-1.05

Ně

-1.04

POSITIVE LOGITS

Ingredienti

1.54

تحميل

1.47

بسم

1.43

がありますが

1.42

Frage

1.41

Zutaten

1.38

آموزش

1.37

Articolo

1.36

Описание

1.30

 faciliter

1.30

Activations Density 0.000%