INDEX

Explanations

geographical and specific nouns

The neuron fires strongly on named entities—proper nouns like organization names, places, and other capitalized multiword terms.

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

 geliefert

-1.23

 שאי

-1.23

 hitherto

-1.21

 pieniądze

-1.21

RIEL

-1.20

 marbre

-1.20

EMBLE

-1.16

 Tutte

-1.16

 Ofrece

-1.16

ETERS

-1.15

POSITIVE LOGITS

’.

1.30

 می‌باشد

1.27

of

1.24

%.

1.22

ENCIÓN

1.19

}*/

1.18

()]

1.16

”.

1.15

brille

1.14

our

1.13

Activations Density 0.066%