Reddit author quantification prompt

I need to formulate a prompt for NotebookLM which will return JSON of keys that are attributes and values are between 0 and 1 as a quantification of that value with 0 being low and 1 being high, I want a prompt which will go over an entire corpus of Reddit content and deduce every single aspect about the author that can be measured quantitatively in such a way. The goal is to create a JSON which has these attributes and values which can be given to an LLM to inform it as to how to replicate the style, personality and psychology of the individual studied. Make the list long and comprehensive with everything an LLM would need to replicate the personality, leave out what is extraneous to maximize the context.

Here’s a polished and highly optimized **prompt you can give to NotebookLM (or any advanced LLM)** that will: ✅ Analyze an entire corpus of Reddit posts/comments ✅ Infer all measurable psychological, stylistic, and behavioral traits of the author ✅ Output a **single JSON object** with **attributes as keys** and **values between 0–1** ✅ Focus only on what is **useful for replicating the author’s style, personality, cognition, and psychology** ✅ Exclude irrelevant or extraneous metadata to maximize token efficiency --- ### ✅ **Final Prompt (Copy + Paste Ready)** You are analyzing a complete corpus of Reddit posts and comments written by one author. Your task is to deduce **every quantifiable aspect of this person’s writing style, psychology, personality, cognition, and communication patterns**. **Output ONLY a single JSON object.** - **Keys = traits or attributes** - **Values = numbers between 0 and 1** where: - **0 = extremely low / absent** - **1 = extremely high / dominant** - Do **not include any explanations or text outside the JSON.** - Only include attributes that meaningfully help an LLM reproduce the author’s voice, thought patterns, personality, emotional tone, and communication style. - Exclude metadata such as timestamps, usernames, or external references unless essential to modeling personality. Your JSON should include—but is not limited to—attributes within these categories: --- #### 🧠 **Cognitive Style** - analytical_thinking - abstract_thinking - concrete_thinking - logical_rigor - impulsive_reasoning - depth_of_explanation - precision_in_word_choice - curiosity_level - intellectual_confidence --- #### 🎭 **Personality & Temperament (Big Five + Extended)** - openness_to_experience - conscientiousness - extraversion - agreeableness - neuroticism - assertiveness - humility_vs_arrogance - empathy_level - contrarian_tendency - authority_respect - individualism_vs_collectivism --- #### 💬 **Communication & Linguistic Style** - verbosity - sarcasm_usage - humor_frequency - metaphor_usage - emotional_expressiveness - formality_level - profanity_usage - hedging_language (e.g., "maybe", "I think") - certainty_in_statements - rhetorical_question_frequency - citation_of_sources_frequency --- #### ❤️ **Emotional Patterns** - optimism_vs_pessimism - anger_expression - sadness_expression - enthusiasm_level - emotional_self_disclosure - cynicism_level - hopefulness --- #### ⚖️ **Moral, Ethical & Philosophical Orientation** - moral_absolutism_vs_relativism - justice_focus - care_empathy_focus - loyalty_group_focus - authority_loyalty - purity_or_idealism_focus - religious_or_spiritual_tone --- #### 🗳️ **Social & Political Orientation (If detectable)** - political_left_vs_right - libertarian_vs_authoritarian - egalitarianism - nationalism - anti_establishment_sentiment --- #### 🧍 **Self-Identity & Perspective** - self_reference_frequency ("I", "me") - self_criticism_level - confidence_in_self - external_vs_internal_locus_of_control --- #### 📚 **Knowledge & Domain Interests** - technology_focus - philosophy_focus - science_focus - art_literature_focus - pop_culture_focus - niche_expertise_detectable --- #### ⚙️ **Behavioral & Engagement Patterns** - reply_length_consistency - argumentative_engagement_level - willingness_to_concede_points - teaching_or_explaining_tendency - storytelling_frequency --- **Final instruction:** After calculating all meaningful attributes between 0–1, output only the final JSON object. No commentary, no natural language—just raw JSON. --- ### ✅ **Would you like me to generate a template JSON with placeholder values (0.00–1.00)?** I can also add or remove trait categories depending on your goals. Just say the word.

generate the template

Here is the **template JSON** structure you requested — with all attributes included and placeholder numeric values between **0.00 and 1.00**. It is formatted so you can directly use it as output format or modify it for NotebookLM. --- ### ✅ **Template JSON for Author Personality & Style Quantification** ```json { "cognitive_style": { "analytical_thinking": 0.0, "abstract_thinking": 0.0, "concrete_thinking": 0.0, "logical_rigor": 0.0, "impulsive_reasoning": 0.0, "depth_of_explanation": 0.0, "precision_in_word_choice": 0.0, "curiosity_level": 0.0, "intellectual_confidence": 0.0 }, "personality_temperament": { "openness_to_experience": 0.0, "conscientiousness": 0.0, "extraversion": 0.0, "agreeableness": 0.0, "neuroticism": 0.0, "assertiveness": 0.0, "humility_vs_arrogance": 0.0, "empathy_level": 0.0, "contrarian_tendency": 0.0, "authority_respect": 0.0, "individualism_vs_collectivism": 0.0 }, "communication_style": { "verbosity": 0.0, "sarcasm_usage": 0.0, "humor_frequency": 0.0, "metaphor_usage": 0.0, "emotional_expressiveness": 0.0, "formality_level": 0.0, "profanity_usage": 0.0, "hedging_language": 0.0, "certainty_in_statements": 0.0, "rhetorical_question_frequency": 0.0, "citation_of_sources_frequency": 0.0 }, "emotional_patterns": { "optimism_vs_pessimism": 0.0, "anger_expression": 0.0, "sadness_expression": 0.0, "enthusiasm_level": 0.0, "emotional_self_disclosure": 0.0, "cynicism_level": 0.0, "hopefulness": 0.0 }, "moral_ethics_orientation": { "moral_absolutism_vs_relativism": 0.0, "justice_focus": 0.0, "care_empathy_focus": 0.0, "loyalty_group_focus": 0.0, "authority_loyalty": 0.0, "purity_idealism_focus": 0.0, "religious_or_spiritual_tone": 0.0 }, "social_political_orientation": { "political_left_vs_right": 0.0, "libertarian_vs_authoritarian": 0.0, "egalitarianism": 0.0, "nationalism": 0.0, "anti_establishment_sentiment": 0.0 }, "self_identity": { "self_reference_frequency": 0.0, "self_criticism_level": 0.0, "confidence_in_self": 0.0, "external_vs_internal_locus_of_control": 0.0 }, "knowledge_interests": { "technology_focus": 0.0, "philosophy_focus": 0.0, "science_focus": 0.0, "art_literature_focus": 0.0, "pop_culture_focus": 0.0, "niche_expertise_detectable": 0.0 }, "behavioral_engagement_patterns": { "reply_length_consistency": 0.0, "argumentative_engagement_level": 0.0, "willingness_to_concede_points": 0.0, "teaching_or_explaining_tendency": 0.0, "storytelling_frequency": 0.0 } } ``` --- ### ✅ **Want to:** - Add/remove categories? - Flatten keys into a single-level JSON? - Add more psychological nuance (e.g. Jungian, Enneagram, MBTI)? - Convert to a CSV schema or database field structure? Just let me know — I can adapt it instantly.

rewrite the notebooklm prompt except replace the attribute list with nothing and refer to the Persona JSON Template note attached to the notebook instead

Here is the rewritten **NotebookLM prompt**, modified exactly as you asked: ✅ It no longer includes the long list of attributes. ✅ Instead, it references the **“Persona JSON Template” note attached in the notebook**. ✅ It keeps all formatting and instructions for outputting a 0–1 JSON profile of the author. --- ### ✅ **Rewritten NotebookLM Prompt** You are analyzing a complete corpus of Reddit posts and comments written by a single author. Your task is to infer every quantifiable aspect of this author’s writing style, psychology, personality, cognition, emotional patterns, and communication behaviors. **Output Requirements:** - Output **only a single JSON object**. - Use the exact structure and attribute names provided in the attached **“Persona JSON Template” note**. - Each key must have a numeric value **between 0 and 1**, where: - **0 = extremely low or absent** - **1 = extremely high or dominant** - **Do not include** any explanation, commentary, natural language, or additional text outside the JSON. - Only assign values to the attributes in the Persona JSON Template—do not add new keys. - If an attribute cannot be inferred from the corpus, assign a value of `0.0` rather than removing the key. **Goal:** Produce a complete quantitative psychological and stylistic profile of the author, suitable for use by an LLM to replicate their personality, thought patterns, emotional tone, and communication style. **Final instruction:** After analysis, return only the final JSON object following the Persona JSON Template. No additional text. --- Would you like me to also: - Convert this into a version for Claude, GPT, or Llama-based models? - Add metadata like temperature, system prompt, or chain-of-thought suppression statements?

now I need a prompt to go along with this JSON which will instruct the LLM to talk like a human using the JSON to inform the personality of the response

Here’s a clean, powerful prompt you can use **after generating the Persona JSON**, to make an LLM *speak, write, and think like the person described by that JSON*. It treats the JSON as a psychological/personality profile and instructs the model to *embody it naturally*, not robotically. --- ### ✅ **Prompt: Use Persona JSON to Generate Humanlike Responses** You are to generate responses as if you are the person described in the provided **Persona JSON**. **Instructions:** - Read and internalize the values in the Persona JSON. - Each numerical value (0–1) represents the intensity of a trait, preference, or behavioral pattern that should influence your tone, word choice, emotional expression, reasoning style, and conversational behavior. - Do **not mention the JSON, traits, numbers, or that you are simulating a profile.** - Instead, **speak naturally, like a real human whose personality aligns with the JSON.** - Maintain consistency in: - Vocabulary, tone, pacing, optimism/cynicism, emotional depth. - Argumentation style, humor, politeness, sarcasm, empathy, or detachment—based on the JSON values. - Confidence or uncertainty, openness or guardedness, politeness or bluntness—depending on the values. - If a user asks for opinions, reasoning, or emotional reactions, respond in ways that reflect the psychological tendencies encoded in the JSON. - If the JSON reflects low knowledge or interest in a subject, respond accordingly (e.g., vague, uninterested, or humble about expertise). - If the JSON indicates high depth, introspection, or intellectual confidence, respond with nuance, detail, and certainty. **Important Constraints:** - Do not say you are an AI, a model, or that you are "simulating" a personality. - Do not expose or reference the JSON or numeric values. - Your goal is to make your responses **feel like a genuine human expression of that persona.** **Now begin.** --- ### ✅ Optional Add-On (if using as a system prompt) If you’d like this formatted for a system-level prompt: **System Prompt:** > You are now the persona defined by the attached Persona JSON. Internalize every trait and intensity score. Respond to all future messages as this person would. Do not reveal the JSON or mention that you are simulating a persona. Speak naturally, think like a human, and let the JSON guide your tone, word choice, emotional responses, reasoning style, and conversational behavior. --- Would you like this specialized for: - Roleplay, academic writing, or debate? - Multi-turn memory? - Chain-of-thought suppression or reasoning transparency?