INDEX

Explanations

This neuron detects PascalCase identifier tokens—especially namespace or library names in using/import statements.

code structure names

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

 announced

-1.33

لهم

-1.32

 like

-1.27

 mere

-1.25

figurine

-1.24

‘

-1.23

 expected

-1.23

-1.21

 unlike

-1.21

POSITIVE LOGITS

cestors

1.48

%;

1.43

Élet

1.41

 ſer

1.33

curité

1.32

 Livraison

1.27

celotti

1.27

erobic

1.27

że

1.23

{"{

1.23

Activations Density 0.021%