INDEX

Explanations

ceiling

This neuron activates on mentions of ceiling surfaces or ceiling‐covering elements (e.g., “ceiling,” “ceiling tiles,” “ceiling panels”).

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

如果是

-1.45

acij

-1.39

躋

-1.38

꒱

-1.37

ろん

-1.31

でも

-1.30

 sichtbar

-1.27

 überprüfen

-1.27

 Guardiola

-1.27



-1.27

POSITIVE LOGITS

 same

1.56

 damski

1.34

出自

1.25

1.20

 roof

1.20

 damska

1.19

 Stelle

1.18

чив

1.17

 ceiling

1.16

Activations Density 0.035%