INDEX

Explanations

xhtml attributes and tags

The neuron is triggered by programming/markup identifiers and keywords—tokens like “align,” “typeof,” or “Location” that name code constructs or environment elements.

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

 utilize

-1.67

has

-1.66

):

-1.59

 allow

-1.58

 well

-1.46

 remain

-1.45

 receive

-1.43

 reichlich

-1.37

觔

-1.35

 ensures

-1.34

POSITIVE LOGITS

petits

1.61

 niektórych

1.59

each

1.59

 своей

1.58

marzo

1.57

 naszych

1.51

pexpr

1.51

emand

1.50

of

1.50

 šed

1.49

Activations Density 0.001%