created by

The neuron fires on the standard multi‐line comment headers at the top of source files (the blocks listing creation dates, copyright notices, and licensing metadata).

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

 romantis

-1.49

 Sécurité

-1.35

camion

-1.30

 verschillende

-1.27

 Bretaña

-1.24

 différentes

-1.23

 Automó

-1.23

 harap

-1.22

 Francus

-1.22

 Privacidad

-1.20

POSITIVE LOGITS

am

1.38

1.27

1.20

1.16

 "{\"

1.16

1.12

was

1.11

Activations Density 0.003%