INDEX

Explanations

packages of items

The neuron detects mentions of packaged quantities—container words like “pack,” “packet,” “bag,” “box,” or “container” when they indicate a specific amount.

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

 pelicula

-1.58

橚

-1.56

⌗

-1.55

sonian

-1.55

 ecran

-1.52

“:

-1.52

"\\

-1.50

"%(

-1.47

anjutan

-1.45

⚣

-1.45

POSITIVE LOGITS

of

2.34

);

1.58

 químicas

1.45

 específicas

1.38

(«

1.37

 explores

1.36

 físicas

1.36

 médicas

1.34

 mobilité

1.34

</h2>

1.34

Activations Density 0.019%