short, topic-defining headings or key nouns that state the main subject of the text or user request.

conversation turn delimiters (especially end-of-turn) and short, title-like user prompts.

code/documentation formatting markers and short identifiers within technical snippets (e.g., bullets, flags, comments, and variable-like tokens).

the start of an assistant’s reply, especially introductory framing that sets up the discussion of the user’s topic.

identifiers and tokens from code snippets, especially snake_case function/variable names with underscores and related code-format elements.

technical syntax and symbol-heavy tokens in code or formatted text, especially XPath expressions like following-sibling and similar structured snippets.

explanatory openings that introduce a step-by-step breakdown of a technical item (e.g., code, commands, or messages).

key section headers and bolded, action-oriented list items in structured, instructional responses.

sexually explicit or suggestive content, including nudity, erotic scenarios, and discussions of sexualization or objectification.

descriptions of artificial intelligence and futuristic science-fiction scenarios, especially space colonization, advanced technology, and formal techno-policy or military-style discourse.

mentions of adolescents—especially explicit teen ages or references to teenage status and context.

references to the fashion modeling world and related industry contexts.

first-person narratives about romantic relationship conflict, especially shifts in affection, betrayal/infidelity, and communication breakdowns.

formatting and structural cues in prompts and dialogues, such as section labels, list items, numbering, and emphasized elements

structured how-to or advice-style breakdowns that explain what to do, when, and why, often organized into clear steps with safety guidance.

scripted multi-speaker dialogue structure, especially speaker turn labels and direct-address conversational turns typical of roleplay or staged conversations.

sentence openings that initiate “how”-type questions, with especially strong response to quantitative formulations.

references to emergency medical response—especially first aid/CPR, rescuing injured or unconscious people, and contacting emergency services.

mentions of female people—she/her subjects, women’s roles or names—especially in intimate, relational, or caregiving contexts.

structural and discourse cues of the model’s step-by-step math solution (e.g., response headers, newlines/section breaks, and procedural lead-ins indicating the start of an explanation).