INDEX

Explanations

owner and authors

The neuron detects mentions of content or property rights holders—terms like “owner,” “owners,” “authors,” “contributors,” and “copyright.”

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

 ferocious

-1.12

effectively

-1.06

 ۱۵

-1.06

 möjligt

-1.05

 filha

-1.02

IFICACIONES

-1.01

 мень

-1.01

 LITER

-0.97

 gigantic

-0.96

épreuve

-0.95

POSITIVE LOGITS

of

3.33

who

2.34

if

1.40

or

1.37

 whose

1.31

勛

1.16

 originali

1.10

ของ

1.09

 whom

1.06

 belangrij

1.05

Activations Density 0.043%