INDEX

Explanations

command and control centers

The neuron detects mentions of organizational command or control centers (e.g., “control,” “command,” “centre,” “headquarters”).

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

 عندما

-1.62

When

-1.55

 всички

-1.54

место

-1.53

 Когда

-1.51

Approximately

-1.50

 เขา

-1.50

 alebo

-1.50

 When

-1.47

entière

-1.47

POSITIVE LOGITS

🌻

1.55

if

1.51

 belga

1.50

⢊

1.49

endon

1.46

 itſelf

1.45

銕

1.45

err

1.44

 cifra

1.40

pez

1.40

Activations Density 0.016%