INDEX

Explanations

places and organizations

The neuron activates most strongly on words denoting places, regions, or communities (e.g., community, regions, country, village, city).

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

who

-1.59

 своих

-1.53

 Tatsache

-1.27

留在

-1.24

 uniquement

-1.21

失去

-1.20

 seulement

-1.16

 setIs

-1.13

Only

-1.13

 zwischen

-1.13

POSITIVE LOGITS

 with

1.41

for

1.24

éter

1.10

or

1.09

 Kemudian

1.09

 then

1.05

 there

1.05

 Resmi

1.05

 setempat

1.02

 μπορεί

1.02

Activations Density 0.061%