INDEX

Explanations

citations and publication details

The neuron strongly activates on numeric strings in scholarly citations (e.g. volume, page, and year numbers in parentheses).

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

ৢ

-1.70

 beziehungs

-1.39

 uLocal

-1.37

icznej

-1.35



-1.34

Configurer

-1.33

Ｎ

-1.30

 repuesto

-1.27

각

-1.27

一

-1.26

POSITIVE LOGITS

 with

1.49

any

1.38

 this

1.36

蹉

1.29

 also

1.29

 myös

1.25

in

1.23

 też

1.22

 these

1.20

理由は

1.20

Activations Density 0.033%