INDEX

Explanations

seeing, lenses, eyes, perspective

The neuron is looking for words that evoke vision or perspective—terms like “lens,” “eyes,” or “glasses.”

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

 gratuitas

-1.26

 öns

-1.24

 souha

-1.21

 Several

-1.20

erweise

-1.20

ַח

-1.17

请选择

-1.15

uyên

-1.15

关键

-1.13

參考

-1.12

POSITIVE LOGITS

();

1.39

 view

1.34

ᴇ

1.29

 ロゴ

1.23

 filter

1.17

to

1.17

herited

1.16

 cittadino

1.16

 vincit

1.15

絎

1.14

Activations Density 0.050%