INDEX

Explanations

organization names or compound nouns

The neuron activates on capitalized tokens that are part of multiword proper names—particularly organization, club, association, and place names.

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

どれも

-1.21

 redonde

-1.14

-1.11

feest

-1.01

is

-0.97

どちらも

-0.96

 وزير

-0.95

 testen

-0.89

 хуже

-0.88

 легче

-0.88

POSITIVE LOGITS

 anaknya

1.22

%，

1.22

大家

1.20

 group

1.10

荩

1.07

cji

1.06

aringan

1.05

 nélk

1.03

baño

1.03

 sangat

1.02

Activations Density 0.060%