INDEX

Explanations

list ending identifiers

The neuron detects the large numeric identifiers (e.g. reference numbers, page/volume codes) used in academic citation entries.

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

get

-1.08

to

-1.02

：……

-1.01

 Also

-1.00

жки

-0.96

 Both

-0.96

≅

-0.96

Are

-0.95

৳

-0.95

 come

-0.94

POSITIVE LOGITS

etc

1.83

อื่น

1.58

and

1.52

 lainnya

1.45

 other

1.44

 autres

1.38

 אחרים

1.30

 jiné

1.23

 dalších

1.20

等は

1.16

Activations Density 0.001%