The neuron strongly activates on unusually rare or infrequent tokens—i.e. words that occur very infrequently in the corpus.

The neuron specifically detects numeric tokens representing decimal or floating-point numbers.

The neuron is highly sensitive to numeric and quantitative expressions (numbers, measurements, counts, or associated units).

This neuron fires on imperative “add” instructions (and similar command verbs) in step-by-step or tutorial language.

The neuron activates on multi‐letter uppercase abbreviations (acronyms), especially when they appear in parentheses.

This neuron strongly fires on the word “all” when it appears in emphatic multi‐word expressions (e.g. “It all happened so quickly,” “Now all in one line”).

The neuron activates on numeric measurements (quantities, units, and values) in the text.

The neuron specializes in spotting technical or scientific terminology related to physical quantities and mathematical expressions (e.g. mentions of angles, currents, torques, equilibria, and related parameters).

This neuron fires most strongly on tokens that single out or refer to a target group or category (e.g. “those,” “patients,” “website visitors,” etc.).

This neuron strongly activates on passages that present structured numeric data or statistical information (e.g. lists of figures, percentages, or counts).

The neuron fires strongly on numeric tokens—especially floating‐point numbers—flagging decimal figures in the text.

This neuron strongly activates on the first meaningful word of new sections or headings (e.g. the leading content token that begins a paragraph, title, or block).

This neuron selectively activates on numeric tokens written with decimal points (i.e. floating-point numbers).

The main thing this neuron does is pick out the unusually informative or high-weight tokens in a passage—words (including numbers, common function words, and key content terms) that carry the most semantic or structural importance.

The neuron fires on auxiliary or modal expressions that convey ability, permission, or potential.

This neuron is triggered by imperative calls to action—short instructional phrases (often addressed to “you”) like “download this image for free,” “find the most recent pictures,” or “write a review and earn.”

The neuron strongly activates on uncommon, low‐frequency tokens—i.e. those rare or “surprising” words or symbols in the text.

The neuron strongly activates on religious/theological discourse and doctrinal exposition.

The neuron strongly activates on mentions of “God” (and closely related theological terms describing His will, wrath, or relationship), effectively detecting explicit references to the deity.

The neuron strongly activates on single tokens that are proper names, technical terms, or acronyms (e.g. brand and organization names, legislation abbreviations, and other named entities).