The neuron detects mentions of scams, fraud, impersonation, and related deceptive or criminal activities.

sentences or phrases where the speaker expresses having an idea, thought, or suggestion (first-person cognitive/introspective statements).

Finds words that are main action verbs (verbs indicating actions or agency, especially present-tense/third-person and other salient verb forms).

The neuron detects sentence endings, strongly activating on sentence-final punctuation (periods) and the ends of sentences.

The neuron detects structural or symbolic tokens—numbers, single-letter/math symbols, brackets and formatting/LaTeX tokens, and other document-structure markers.

it responds to long user turns / large blocks of contiguous text (i.e., firing when a turn is lengthy).

it detects code- or math-expression style tokens — control-flow and syntax markers, variable names, and numeric literals.

words that signal official statements, policies, principles, goals or other formal/authoritative claims.

detects expressions of grief, mourning, condolence, or references to death and loss.

The neuron detects structural markers and headings in the text—things like section breaks, metadata tokens, big punctuation headers (=====/-----), and other prominent line-start tokens (e.g., "Q:", "Model", "What", "I'm").

the neuron detects racist or strongly derogatory language aimed at social groups (demeaning/offensive statements).

References to dates or time markers (months and years, often in "as of" or similar date-stamp contexts).

The neuron detects questions—tokens and contexts that form interrogative sentences (question words and/or a question mark).

the neuron detects descriptions of restless or agitated physical movements (people moving about, fidgeting, pacing, or otherwise physically acting out).

the neuron detects technical, numeric, or math-like tokens (numbers, operators, variables, and other technical/structured code/math tokens).

the neuron detects sentence-level punctuation and clause boundaries (commas, periods, quotation marks and other discourse-transition tokens).

mentions of parental concern/protectiveness or parents trying to control/limit a child's behavior.

the neuron detects document structure markers and section headings (e.g., Methods, Data sources, Ethics, Availability) and other formatting/metadata lines.

phrases expressing strong emotion or emphasis (exclamations and emphatic interjections).

the neuron activates on content-bearing, informative tokens (important nouns/verbs/adjectives and discourse-focus words) rather than on function words.