INDEX

Explanations

presence of

The neuron activates on mentions of specific technical/scientific entities—particularly specialized materials, compounds, or phenomena (e.g. “dielectrics,” “micelles,” “vasoactive peptides”).

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

 what

-1.63

how

-1.45

of

-1.36

 will

-1.34

 September

-1.32

-1.29

Бележки

-1.27

 these

-1.23

 this

-1.21

return

-1.20

POSITIVE LOGITS

in

1.45

 kré

1.35

klima

1.27

adalah

1.23

 perfekte

1.22

 nossos

1.21

lindung

1.21

 queridos

1.20

 blessure

1.18

 nosso

1.17

Activations Density 0.268%