What is DeepSeek? AI + Domestic + Free + Open Source + Powerful DeepSeek is a Chinese tech company specializing in General Artificial Intelligence (AGI), focusing on the development and application of large models. DeepSeek-R1 is its open-source reasoning model, excelling in handling complex tasks and available for free commercial use. DeepSeek from entry to mastery (Tsinghua University) PDF Downlod   What Can DeepSeek Do? Text Generation Structured Generation: Tables, lists (e.g., schedules, recipes) Document Writing: Code comments, documentation Creative Writing: Articles, stories, poetry, marketing copy, social media content, scripts, etc. Summarization and Rewriting: Long text summaries (papers, reports), text simplification, multilingual translation and localization   Natural Language Understanding and Analysis Knowledge Reasoning: Logical problem-solving (math, common sense reasoning), causal analysis (event correlation) Semantic Analysis: Sentiment analysis (reviews, feedback), intent recognition (customer service, user queries), entity extraction (names, locations, events) Text Classification: Topic labeling (e.g., news categorization), spam content detection   Programming and Code-Related Tasks Code Generation and Completion: Code snippets (Python, JavaScript), auto-completion with comments Code Debugging: Error analysis and repair suggestions, performance optimization tips Technical Documentation: API documentation, codebase explanation and example generation   Conventional Drawing (Not explicitly mentioned but implied through general capabilities)   How to Use DeepSeek? Access:https://chat.deepseek.com From Beginner to Master: When everyone can use AI, how can you use it better and more effectively? Reasoning Models Example: DeepSeek-R1, GPT-3 excel in logical reasoning, mathematical reasoning, and real-time problem-solving. Reasoning models are models that enhance reasoning, logical analysis, and decision-making capabilities on top of traditional large language models. They often incorporate additional technologies such as reinforcement learning, neuro-symbolic reasoning, and meta-learning to strengthen their reasoning and problem-solving abilities. Non-reasoning models are suitable for most tasks. General models typically focus on language generation, context understanding, and natural language processing, without emphasizing deep reasoning capabilities. These models usually grasp language patterns through extensive text data training and can generate appropriate content, but they lack the complex reasoning and decision-making abilities of reasoning models.   Dimension Comparison Dimension Reasoning Model General Model Strengths Mathematical derivation, logical analysis, code generation, complex problem decomposition Text generation, creative writing, multi-turn dialogue, open-ended questions Weaknesses Divergent tasks (e.g., poetry creation) Tasks requiring strict logical chains (e.g., mathematical proofs) Performance Essence Specializes in tasks with high logical density Excels in tasks with high diversity Strength Judgment Not universally stronger, but significantly better in their training target domains More flexible in general scenarios, but requires prompt engineering to compensate for capabilities Example Models GPT-3, GPT-4 (OpenAI), BERT (Google): Mainly used for language generation, language understanding, text classification, translation, etc.     Fast Thinking vs. Slow Thinking Fast Reaction Models (e.g., ChatGPT-4): Quick response, low computational cost, based on probability prediction through extensive data training Slow Thinking Models (e.g., OpenAI-1): Slow response, high computational cost, based on chain-of-thought reasoning to solve problems step-by-step Decision-Making: Fast reaction models rely on pre-set algorithms and rules, while slow thinking models can make autonomous decisions based on real-time analysis Creativity: Fast reaction models are limited to pattern recognition and optimization, while slow thinking models can generate new ideas and solutions Human Interaction: Fast reaction models follow pre-set scripts and struggle with human emotions and intentions, while slow thinking models can interact more naturally and understand complex emotions and intentions Problem-Solving: Fast reaction models excel in structured and well-defined problems, while slow thinking models can handle multi-dimensional and unstructured problems, providing creative solutions Ethical Issues: Fast reaction models as controlled tools have minimal ethical concerns, while slow thinking models raise discussions on autonomy and control   CoT Chain-of-Thought The emergence of CoT chain-of-thought divides large models into two categories: "probability prediction (fast reaction)" models and "chain-of-thought (slow thinking)" models. The former is suitable for quick feedback and immediate tasks, while the latter solves complex problems through reasoning. Understanding their differences helps in choosing the appropriate model for the task to achieve the best results. Prompt Strategy Differences Reasoning Models: Prompts should be concise, focusing directly on the task goal and requirements (as reasoning logic is internalized). Avoid step-by-step guidance, as it may limit the model's capabilities. General Models: Prompts need to explicitly guide reasoning steps (e.g., through CoT prompts), otherwise, the model may skip key logic. Rely on prompt engineering to compensate for capability shortcomings.   Key Principles Model Selection: Choose based on task type, not model popularity (e.g., reasoning models for math tasks, general models for creative tasks). Prompt Design: Reasoning Models: Use concise instructions, focus on the goal, and trust the model's internalized reasoning capabilities. ("Just say what you want.") General Models: Use structured and compensatory guidance. ("Fill in what's missing.")   Avoid Pitfalls: Do not use heuristic prompts (e.g., role-playing) with reasoning models, as it may interfere with their logical mainline. Do not over-trust general models (e.g., directly asking complex reasoning questions); instead, validate results step-by-step.   From "Giving Instructions" to "Expressing Needs" Strategy Types Strategy Type Definition & Goal Applicable Scenarios Example (for Reasoning Models) Advantages & Risks Instruction-Driven Directly provide clear steps or format requirements Simple tasks, quick execution "Write a quicksort function in Python with comments." ✅ Precise and efficient results ✕ Limits model's optimization space Demand-Oriented Describe problem background and goals, let the model plan the solution path Complex problems, model's autonomous reasoning "Optimize the user login process by analyzing current bottlenecks and proposing 3 solutions." ✅ Stimulates model's deep reasoning ✕ Need to clearly define demand boundaries Hybrid Mode Combine problem description with key constraints Balance flexibility and controllability "Design a 3-day travel plan for Hangzhou, including West Lake and Lingyin Temple, with a budget of 2000 yuan." ✅ Balances goals and details ✕ Avoid over-constraining Heuristic Questioning Guide the model to think actively through questions (e.g., "why," "how") Exploratory problems, model's explanatory logic "Why choose gradient descent for this optimization problem? Compare with other algorithms." ✅ Triggers model's self-explanation ability ✕ May deviate from core goals Task Demand and Prompt Strategy Task Type Applicable Model Prompt Focus Example (Effective Prompt) Prompts to Avoid Mathematical Proof Reasoning Model Direct questioning, no step-by-step guidance "Prove the Pythagorean theorem" Redundant decomposition (e.g., "First draw a diagram, then list formulas") Creative Writing Reasoning Model Encourage divergence, set roles/styles "Write an adventure story in Hemingway's style" Over-constraining logic (e.g., "List steps in chronological order") Code Generation Reasoning Model Concise needs, trust model logic "Implement quicksort in Python" Step-by-step guidance (e.g., "First write the recursive function") Multi-turn Dialogue General Model Natural interaction, no structured instructions "What do you think about the future of artificial intelligence?" Forced logical chains (e.g., "Answer in three points") Logical Analysis Reasoning Model Directly pose complex problems "Analyze the utilitarianism and deontological conflict in the trolley problem" Adding subjective guidance (e.g., "Which do you think is better?") General Model General Model Break down problems, ask step-by-step "First explain the trolley problem, then compare the two ethical views" One-time questioning of complex logic How to Express Needs to AI Demand Type Characteristics Demand Expression Formula Reasoning Model Adaptation Strategy General Model Adaptation Strategy Decision-Making Need to weigh options, assess risks, choose the best solution Goal + Options + Evaluation Criteria Request logical deduction and quantitative analysis Direct suggestions, rely on model's experience Analytical Need to deeply understand data/information, discover patterns or causal relationships Problem + Data/Information + Analysis Method Trigger causal chain deduction and hypothesis validation Surface summarization or classification Creative Need to generate novel content (text/design/solution) Theme + Style/Constraints + Innovation Direction Combine logical framework to generate structured creativity Free association, rely on example guidance Verification Need to check logical consistency, data reliability, or solution feasibility Conclusion/Solution + Verification Method + Risk Points Design verification path independently and identify contradictions Simple confirmation, lack of deep deduction Execution Need to complete specific operations (code/calculation/process) Task + Step Constraints + Output Format Optimize steps autonomously, balance efficiency and correctness Strictly follow instructions, no autonomous optimization Prompt Examples Decision-Making Demand: "Two options are available to reduce logistics costs: ① Build a regional warehouse (high initial investment, low long-term costs) ② Partner with a third party (pay-as-you-go, high flexibility) Please use the ROI calculation model to compare the total costs over 5 years and recommend the optimal solution." Verification Demand: "Here is a conclusion from a paper: 'Neural network model A is superior to traditional method B.' Please verify: ① Whether the experimental data supports this conclusion; ② Check if there is any bias in the control group setup; ③ Recalculate the p-value and determine significance." Analytical Demand: "Analyze the sales data of new energy vehicles over the past three years (attached CSV), and explain: ① The correlation between growth trends and policy; ② Predict the market share in 2025 using the ARIMA model and explain the basis for parameter selection." Execution Demand: "Convert the following C code to Python, with the following requirements: ① Maintain the same time complexity; ② Use numpy to optimize array operations; ③ Output the complete code with time test cases." Creative Demand: "Design a smart home product to address the safety issues of elderly people living alone, combining sensor networks and AI early warning. Provide three different technical route prototype sketches with explanations." Verification Demand: "Below is a conclusion from a paper: 'Neural network model A is superior to traditional method B.' Please verify: ① Whether the experimental data supports this conclusion; ② Check if there is any bias in the control group setup; ③ Recalculate the p-value and determine significance." Do We Still Need to Learn Prompts? Prompts are the instructions or information that users input into an AI system to guide it to generate specific outputs or perform specific tasks. Simply put, prompts are the language we use to "converse" with AI. They can be a simple question, a detailed instruction, or a complex task description. A prompt consists of three basic elements: Instruction (Instruction): The core of the prompt, explicitly telling the AI what task to perform. Context (Context): Providing background information to help the AI better understand and execute the task. Expectation (Expectation): Clearly or implicitly expressing the requirements and expectations for the AI's output.   Types of Prompts Instructional Prompts: Directly tell the AI what task to perform. Question-Answer Prompts: Pose questions to the AI, expecting corresponding answers. Role-Playing Prompts: Require the AI to assume a specific role and simulate a particular scenario. Creative Prompts: Guide the AI to perform creative writing or content generation. Analytical Prompts: Require the AI to analyze and reason about given information. Multimodal Prompts: Combine text, images, and other forms of input.   The Essence of Prompts Feature Description Example Communication Bridge Connects human intent with AI understanding "Translate the following into French: Hello, world" Context Provider Provides necessary background information for the AI "Assuming you are a 19th-century historian, comment on Napoleon's rise" Task Definer Clearly specifies the task the AI needs to complete "Write an introduction for an article on climate change, 200 words" Output Shaper Influences the form and content of the AI's output "Explain quantum mechanics in simple terms, as if speaking to a 10-year-old" AI Capability Guide Guides the AI to use specific abilities or skills "Use your creative writing skills to create a short story about time travel" Article from:Team: Yu Menglong, Postdoctoral Fellow Tsinghua University School of Journalism and Communication New Media Research Center, Metaverse Culture Lab