INDEX

Explanations

code identifiers

This neuron fires on capitalized code identifiers—especially type or component names (like class, interface, or prop names) in source files.

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

and

-2.06

 cũng

-1.76

of

-1.55

KHz

-1.48

 disheart

-1.47

now

-1.46

id

-1.45

 also

-1.45

}$

-1.43

 distraught

-1.43

POSITIVE LOGITS

 gustaba

1.74

 redenen

1.65

![](

1.53

 hubieran

1.45

 fresas

1.44

 rempliss

1.40

 tendremos

1.40

 garansi

1.38

鏨

1.38

 แล้ว

1.37

Activations Density 0.078%