INDEX

Explanations

assigning values to attributes

The neuron fires on parameter or attribute names (like “color,” “link,” “scale,” “image,” “Tag,” etc.) used in key–value settings across code and markup.

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

 will

-1.46

 with

-1.45

無し

-1.41

為に

-1.41

 this

-1.38

 enough

-1.34

 that

-1.31

({

-1.30

 which

-1.27

に入れ

-1.27

POSITIVE LOGITS

や

1.63

などを

1.53

がなく

1.50

を

1.46

ftet

1.45

は

1.44

<bos>

1.39

Biblia

1.38

Manfaat

1.37

Ofic

1.35

Activations Density 0.001%