INDEX

Explanations

numbers followed by )

The neuron selectively fires on standalone numeric tokens that correspond to legal citation elements (years, volume numbers, and page numbers).

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

of

-1.70

={

-1.27

-1.22

-1.20

-1.10

 getchar

-1.09

 kutoka

-1.07

 conosciuto

-1.06

 Research

-1.05

=[];

-1.05

POSITIVE LOGITS

Sziasztok

1.31

阳台

1.26

 gelatina

1.23

🟡

1.23

ּוֹ

1.22

前は

1.22

がありますが

1.21

แค่

1.20

 vitamina

1.20

Когда

1.17

Activations Density 0.033%