INDEX

Explanations

URLs or network addresses

This neuron detects numeric literals (decimal or integer) in XML attribute values.

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

not

-1.59

สำหรับ

-1.41

➪

-1.37

土豆

-1.35

<tbody>

-1.33

out

-1.33

烈的

-1.32

 για

-1.30

合わせた

-1.30

 untuk

-1.24

POSITIVE LOGITS

 ANOTHER

1.84

 condomínio

1.81

釹

1.67

 artesanato

1.60

 prédio

1.57

 caneca

1.54

 escritório

1.52

 あり

1.52

 Überprüfung

1.50

 créé

1.49

Activations Density 0.034%