INDEX

Explanations

adopted as

The neuron fires on words about taking something on or appropriating it—especially “adopted” (and variants like “co-opted”).

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

 информация

-1.28

 entret

-1.27

-1.23

 valenci

-1.21

实力

-1.17

 тыся

-1.15

-1.14

 PRESIDENT

-1.12

会儿

-1.11

 Schwer

-1.11

POSITIVE LOGITS

 Particularly

1.42

for

1.41

 sogenannten

1.34

 Для

1.30

 especially

1.30

 ваго

1.26

 damaligen

1.23

厦

1.23

 Bildern

1.22

 allgemeinen

1.22

Activations Density 0.032%