INDEX

Explanations

private member declarations

The neuron consistently lights up on code tokens that control or modify access to data—e.g. visibility keywords like “private,” property‐accessors like “set,” “readonly,” and assignment/arrow operators.

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

葙

-1.76

you

-1.63

 Süß

-1.50

 varias

-1.45

ACKNOWLEDGMENT

-1.45

を見

-1.43

なんでも

-1.42

Sicherheit

-1.41

籏

-1.41

 maybe

-1.40

POSITIVE LOGITS

𓄹

1.88

ном

1.76

and

1.73

).

1.65

;"

1.63

ühle

1.56

But

1.52

 estekak

1.51

"--

1.50

 videojuegos

1.47

Activations Density 0.001%