INDEX

Explanations

contested estimates or evolving practices

This neuron detects formal academic or technical jargon—long, multisyllabic scholarly terms and discourse markers typical of academic writing.

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

 inizi

-1.35

 telles

-1.30

 such

-1.30

 otwor

-1.26

xFFFF

-1.23

 ????

-1.23

乄

-1.19

-1.18

地區

-1.16

xA

-1.15

POSITIVE LOGITS

นอก

1.23

 chút

1.23

astră

1.22

ιώ

1.16

क्या

1.15

 感じ

1.15

끓

1.15

˩

1.15

boBox

1.15

 unsafe

1.15

Activations Density 0.154%