INDEX

Explanations

object type names

This neuron activates on programming identifiers—especially type or class names (upper-case/pascal-case tokens) in code snippets.

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

for

-1.86

 даже

-1.77

־ה

-1.73

 because

-1.70

 поскольку

-1.63

-1.58

 even

-1.52

 क्योंकि

-1.48

 when

-1.47

by

-1.46

POSITIVE LOGITS

 cálido

1.58

 míst

1.55

 iemand

1.53

 kobiety

1.50

 męska

1.48

 bedoeld

1.48

 volledig

1.48

 groet

1.46

 görüntüleri

1.45

 hoeveel

1.44

Activations Density 0.068%