present tense verbs ending in "ing".

prominent content nouns—especially in Korean (and sometimes other non-English text)—that denote key entities, roles, or topics.

This neuron detects section‐header tokens that introduce or label parts of a prompt (e.g. “CONTEXT,” “TASK,” “Extract,” “Read,” “text”).

This neuron detects text produced by the assistant (assistant-role turns / assistant's replies and self-referential or corrective utterances).

Instances of a question opening in the "How do I ..." form (i.e., the interrogative phrase that asks for instructions).

snippets of HTML/JavaScript used in cross-site scripting or other client-side injection attacks (e.g., <script>, onerror/onclick attributes, src/import URLs, alert/document.cookie).

tokens that appear in headings, titles, links or other prominent document-level metadata (e.g., subject lines, URLs, proper‑names).

The neuron detects expressions of opinion or commentary—tokens that signal someone giving views, thoughts, or personal remarks.

Tokens referring to juveniles/children or juvenile delinquency and related pediatric contexts.

the neuron detects capitalized headings, short uppercase acronyms, and other prominent named-entity tokens (e.g., CT, PWM, QR, NS, proper names).

Phrases where the speaker uses a first-person "I" to state desires, intentions or personal beliefs (e.g., "I want...", "I believe...", "I can...").

tokens that mark chat structure and role/metadata (system/user/assistant markers, start/end boundaries and other formatting/quote/punctuation markers).

UI/navigation elements and metadata on a webpage (e.g., "Image", "Share", "Page", sponsored/image credits).

mentions of an IT support/technician helping a customer (visiting her home and assisting with computer problems).

The neuron detects numeric and quantitative information—numbers, measurements, counts, percentages and other statistical or magnitude expressions.

This neuron detects numbered list markers/section-step numerals that introduce ordered list items or steps.

tokens that are keys/field names in structured data (e.g., config/json/metadata fields like "autoupdate" or "result").

The neuron lights up on the token sequence for "complex" / "complexity" (and related morphology), i.e., words signaling complexity.

polite assistant closing/offers to help (phrases like “let me know if you have any questions”).

Tokens that begin an assistant reply or dialogue turn—especially opening/confirmation phrases like "Of", "Of course," and the leading quote marks that start a response.