INDEX

Explanations

program code elements

The neuron activates on references to numeric‐type classes and their conversion methods (e.g. Integer.parseInt, Long.class, Float.parseFloat, Convert.ToString).

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

！"

-1.57

回し

-1.50

しました

-1.48

𓐍

-1.45

 wich

-1.45

焢

-1.45

 lavable

-1.39

-1.38

 voiture

-1.38

?"

-1.38

POSITIVE LOGITS

to

1.69

edere

1.55

UPAC

1.52

的重要性

1.41

 zugänglich

1.40

ನ

1.38

ANAL

1.37

 nachfol

1.36

 wysokość

1.34

});

1.30

Activations Density 0.003%