INDEX

Explanations

mathematical variable X

The neuron selectively fires on uppercase mathematical variables—especially “X” (and similarly “A,” “B,” etc.) used as matrix or linear‐algebra symbols.

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

or

-1.72

-1.63

］

-1.48

]-->

-1.45

 after

-1.44

-1.41

}}$

-1.37

timezone

-1.36

));

-1.35

in

-1.34

POSITIVE LOGITS

 siè

1.45

︲

1.42

 sechs

1.41

試試

1.39

 verschillende

1.39

 zwölf

1.38

摻

1.36

 ihrem

1.35

 กับ

1.34



1.34

Activations Density 0.095%