INDEX

Explanations

series of

This neuron detects section headings that specify a “Series” number followed by a Roman‐numeral or colon label (e.g. “Series 328 I:” or “series 320”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

orphism

-2.03

ﱠ

-1.88

欹

-1.81

ейчас

-1.72

ارتمان

-1.70

ıyordu

-1.69

向けの

-1.66

泚

-1.66

 solicitudes

-1.63

if

-1.62

POSITIVE LOGITS

of

2.11

.”

1.99

 серии

1.88

1.86

 serisi

1.80

1.72

ของ

1.72

 anderen

1.68

 SERIES

1.67

 Series

1.65

Activations Density 0.056%