INDEX

Explanations

dimensional / internal concepts

The neuron detects dense theoretical‐physics jargon and formulaic tokens—things like gauge theories, extra dimensions, manifolds, AdS notation, and other specialized technical terms.

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

 kunnen

-1.50

 crea

-1.49

 especific

-1.44

 separa

-1.37

 conclusión

-1.34

 said

-1.33

értelmű

-1.30

 propias

-1.28

とその

-1.28

 만들

-1.27

POSITIVE LOGITS

</strong>

1.58

 with

1.52

1.47

 Already

1.45

</h3>

1.33

NDICE

1.27

any

1.24

<h3>

1.24

 Eleven

1.23

up

1.23

Activations Density 0.019%