INDEX

Explanations

citations and references

This neuron detects tokens that are part of scholarly reference entries—author names and initials in citation lists.

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

Ⅱ

-1.50

nargs

-1.48

 alkoh

-1.40

==$

-1.39

 {}));

-1.36

'),

-1.31

maxAge

-1.31

𝐃

-1.27

<0xF3>

-1.27

 Canadá

-1.27

POSITIVE LOGITS

了不少

1.37

了很多

1.36

 diversas

1.32

 verschiedenen

1.31

 diversos

1.30

难怪

1.28

 verschiedener

1.27

 विभिन्न

1.27

 dismant

1.26

了许多

1.26

Activations Density 0.003%