INDEX

Explanations

This neuron specializes in spotting specific entity nouns—names of people, places, organizations, or defined groups.

places and roles

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

ほら

-1.46

つづく

-1.33

ようで

-1.30

うわ

-1.30

 deaktiviert

-1.23

おそ

-1.22

になるので

-1.21

えっ

-1.20

 créée

-1.18

為に

-1.18

POSITIVE LOGITS

 came

1.41

 nonché

1.38

これを

1.37

 have

1.29

 wrote

1.24

 went

1.21

 finally

1.19

 took

1.18

その

1.16

 make

1.15

Activations Density 0.072%