INDEX

Explanations

class and namespace declarations

The neuron detects high-level code‐structure keywords that introduce or define classes, methods, or namespaces (e.g. “public,” “class,” “void,” “main,” “using,” “namespace”).

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

 recientemente

-1.55

了个

-1.46

 citada

-1.39

 cumbre

-1.36

了自己的

-1.36

着自己的

-1.34

ffindor

-1.34

 sogen

-1.32

了几

-1.31

 PROPER

-1.30

POSITIVE LOGITS

.”

1.80

 will

1.65

》。

1.58

。

1.57

?”

1.56

 that

1.54

мі

1.52

”.

1.50

%.

1.50

</em>

1.48

Activations Density 0.050%