INDEX

Explanations

menu headings and foreign terms

The neuron selectively activates on standalone numeric tokens (i.e. numbers written as digits).

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

-1.48

in

-1.30

 that

-1.22

 first

-1.11

$.}

-1.09

-1.06

 utford

-1.05



-1.02



-1.02

-1.01

POSITIVE LOGITS

and

1.37

menü

1.22

 prachtige

1.20

gespräch

1.20

šenie

1.13

 algunos

1.13

 konusu

1.09

 şeyi

1.08

 varier

1.06

bestimmungen

1.06

Activations Density 0.000%