INDEX

Explanations

quoting and assignment delimiters

The neuron is specialized to pick up numeric expressions—tokens representing numbers, measurements, durations, currency amounts, or other digit‐containing values.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits



0.59

。（

0.57

↵↵↵↵↵

0.56



0.52

’?

0.51

‘

0.51

↵↵↵↵↵↵

0.50

Moral

0.49

’),

0.49

relatively

0.49

POSITIVE LOGITS

\""

1.81

'"

1.53

=\""

1.37

='"

1.33

"+

1.27

"'

1.23

:"+

1.23

#"

1.23

='"+

1.20

-"+

1.17

Activations Density 0.009%