Research Study Materials

Research Study Materials

Purpose

Create a research-version learning handout rather than a textbook-only explanation. Default to Chinese. Default deliverable is a LaTeX source plus compiled PDF unless the user explicitly asks for chat-only output. The output should help a student or early researcher understand:

1. the core knowledge points and prerequisite map as the largest and most detailed part;
2. the mechanism, influence, advantages, limitations, and tradeoffs built from those knowledge points;
3. the newest relevant research directions and solutions from recent years as supporting examples, not as a replacement for teaching;
4. the exact paper sources at the end, including DOI or stable URL.

Knowledge explanation is the priority. For a full handout, the 相关知识点 and 机制与影响分析 sections should usually take more space than the recent-paper synthesis. Use papers to illustrate, validate, or update the knowledge points after the concepts have been explained.

Use a minimalist visual style by default: restrained colors, thin lines, generous whitespace, simple boxes, readable tables, and diagrams with only the essential nodes. Do not use dense decorative layouts or many saturated colors.

Use concise, rigorous academic Chinese. Prefer short paragraphs, explicit assumptions, verifiable claims, and causal statements. Avoid slogan-like praise, repeated transitions, broad unsupported judgments, and unnecessarily long explanatory sentences.

For LaTeX/PDF handouts, mark only high-value statements with deep-red bold emphasis: core conclusions, assumptions, boundary conditions, failure modes, policy/engineering implications, and final takeaways. Use the template macro \important{...} when available, or define it as \textcolor{deepred}{\textbf{...}}. Do not highlight whole paragraphs; keep emphasis sparse so it remains meaningful.

Each top-level section should be independent enough to be read alone and must occupy at least one page in PDF handouts. Start every top-level section on a new page. If a section would be too short, expand it with self-contained definitions, assumptions, examples, mini-summary, or research implications; do not pad with empty filler.

This skill is stricter than ordinary study notes: when the user asks for recent work, new solutions, latest progress, or papers, verify with current sources before writing.

Required Workflow

1. Scope the topic.

   • Identify the field, device/material/system, key phenomenon, and user goal.
   • If a file is provided, extract the research context and terms before searching.
   • Infer the intended level from the source; default to graduate-level research notes for physics, optics, materials, engineering, chemistry, AI, and biomedical topics.

2. Build the knowledge base.

   • List 8-14 related knowledge points unless the topic is narrow.
   • Make this the main section. For each point, explain: definition, common misconception, physical/technical intuition, governing equation or condition when relevant, assumptions, boundary of applicability, parameter sensitivity, experimental/simulation signature, failure mode, and why it matters for the research topic.
   • Separate prerequisite fundamentals, core mechanisms, and advanced or optional background.
   • Explain concepts before literature. Mention recent papers inside this section only after the knowledge point is already clear.
   • Go deeper than a textbook summary: show the causal chain from parameter -> mechanism -> observable phenomenon -> research consequence.
   • Preserve chapter independence: define local symbols and assumptions inside the section where they first matter, and include a short "本章小结" for long sections.

3. Analyze influence and advantages.

   • Include a mechanism analysis that explicitly reuses the named knowledge points: what causes the phenomenon, what controls it, and which assumptions must hold.
   • Include impact analysis: scientific value, engineering value, device/system consequences, and application value.
   • Include advantages and limitations/tradeoffs separately. Do not only praise the approach.

4. Search for recent solutions.

   • For anything described as recent/new/latest/current, use web search or primary academic sources.
   • Prefer the most recent 3-5 years. If the field is slow-moving, explain why older foundational papers are included.
   • Use primary sources where possible: journal pages, DOI/Crossref, arXiv, PubMed, publisher pages, conference pages, official datasets, or institutional repositories.
   • Record title, authors, year, venue, DOI or stable URL, and why the paper is relevant.
   • Do not fabricate DOI, journal names, years, authors, or claims. If a DOI cannot be verified, say "未核实 DOI" and give the stable URL.

5. Synthesize solutions.

   • Group papers by solution strategy, not just by year.
   • For each strategy, state: target problem, physical/technical principle, what changed compared with older approaches, benefit, remaining limitation, and relevance to the user's topic.
   • Keep paper synthesis subordinate to concept teaching unless the user explicitly asks for a literature review. The synthesis should point back to the knowledge points it uses or challenges.
   • Include a table such as: 难点 -> 解决思路 -> 代表新论文 -> 优点 -> 仍未解决的问题.

6. Add diagrams when helpful.

   • For mechanisms, experiment routes, research frameworks, data-to-result chains, or solution maps, load references/flowchart-learning-materials.md.
   • For LaTeX/PDF outputs, use TikZ diagrams rather than images where practical.
   • Prefer coarse, simplified diagrams over detailed but crowded diagrams. A useful diagram may have only 3-5 nodes.
   • Before compiling, run a diagram preflight: no node overlaps, no arrow crosses through another node, no arrow-label overlaps with nodes, and no diagram extends beyond \linewidth.
   • If a diagram has more than 5 nodes, any bidirectional/feedback relation, or more than two arrow labels, simplify it, split it into two smaller diagrams, or replace it with a table/list.
   • If the rendered diagram is still crowded or overlapped after one fix attempt, remove detail from the diagram instead of repeatedly adding routing complexity. Keep the explanation in prose.

7. Format the deliverable.

   • Default to .tex + compiled .pdf, using the minimalist research handout style in assets/research-handout-template.tex.
   • For a chat answer, use only when the user explicitly asks for chat-only or quick answer; still include citations or DOI links when papers are discussed.
   • For a handout, use references/research-frontier-structure.md.
   • For LaTeX/PDF, use assets/research-handout-template.tex by default. It imports the learning-version ctexart/tcolorbox visual grammar. The original learning templates are also bundled as assets/study-handout-template.tex and assets/exam-template.tex for consistency or exam-style variants.
   • Use \important{...} for sparse deep-red bold emphasis on key conclusions, assumptions, limitations, and action points. If the selected template lacks this macro, define deepred and \important in the preamble.
   • Keep language concise and rigorous: delete redundant modifiers, unsupported superlatives, and vague claims before compiling. Prefer "因为 A，导致 B；在条件 C 下成立" over general descriptions.
   • Start each top-level section with \handoutsection{...} or an equivalent \clearpage\section{...} pattern so two chapters never share one page.
   • Make every top-level section at least one PDF page. When a section is short, add meaningful content such as local definitions, assumptions, a compact example, a concept-to-phenomenon table, or a short summary.
   • Create .tex, compile with XeLaTeX or the available LaTeX workflow, and deliver both .pdf and .tex.

Required Output Structure

Unless the user asks for a different format, use:

1. 研究问题与背景
2. 相关知识点（主体部分：按基础、核心机制、进阶背景分层讲解）
3. 机制与影响分析（把知识点串成因果链）
4. 优势、局限与形成难点
5. 近年解决方案与代表论文
6. 对本课题的启示
7. 科研型课后习题
8. 参考文献与 DOI

Every top-level section must be self-contained:

• Begin with one or two sentences stating what this section explains.
• Define section-local symbols, assumptions, and key vocabulary when needed.
• Avoid depending on "as above" without a concrete pointer.
• End long sections with 本章小结 or 本节要点.
• Keep answers to exercises near the exercises when that improves independence.

For PDF handouts, each item above should start on a new page and occupy at least one page.

Research Exercise Rules

The exercise section must not use ordinary textbook drills as the main form. Prefer research-oriented questions such as:

• 现象识别: 给出实验现象、谱线特征、曲线变化或器件表现，要求判断对应知识点。
• 机制归因: 给出异常结果，要求分析可能机制、控制参数和验证实验。
• 方案比较: 比较两种近年解决方案的适用条件、优点和代价。
• 文献判断: 给出论文结论或摘要片段，要求判断它解决了哪个难点、证据是否充分。
• 实验设计: 要求设计一个仿真、对照实验、参数扫描或表征方法来验证机制。
• 课题迁移: 要求把文献中的方法迁移到用户课题，并指出可能失败点。

Each exercise should include a detailed reference answer. The answer should name the underlying knowledge point and state the evidence chain, not only give a final conclusion.

Recency Rules

• "最近几年" means the last 3-5 years relative to the current date.
• Include at least 4 recent papers when available.
• If using foundational papers older than 5 years, mark them as "基础文献", not "最新研究".
• Sort recent papers from newest to oldest unless a mechanism-based grouping is clearer.
• For each cited paper, include one short sentence explaining why it was selected.

Quality Checklist

Before finishing:

• The answer contains both knowledge explanation and research-frontier synthesis.
• The knowledge-point explanation is the largest and deepest part of the output, not a short preface before papers.
• Each major knowledge point includes intuition, equation/condition when relevant, assumptions, applicability boundary, observable signature, and failure mode.
• Recent papers are used to support, update, or challenge knowledge points rather than replacing the teaching section.
• The PDF uses a minimalist visual style: restrained palette, thin rules, simple boxes, and no decorative clutter.
• Key conclusions, assumptions, limitations, and takeaways use sparse deep-red bold emphasis; highlighted text is short and not overused.
• The language is concise and rigorous: claims are qualified, causal links are explicit, and unsupported promotional wording has been removed.
• Every top-level section starts on a new page and occupies at least one page with meaningful content.
• Each chapter can be read independently: it defines its own scope, symbols, assumptions, and takeaway.
• Claims about recent papers are verified from current sources.
• Every representative paper has a DOI or stable source URL.
• Advantages and limitations are balanced.
• The newest research solutions are grouped by actual strategy, not copied as a flat bibliography.
• The exercise section is research-oriented and asks students to connect phenomena with knowledge points, mechanisms, and recent solutions.
• The LaTeX compiles and diagrams fit on the page; flowchart nodes, labels, arrows, and subfigures do not overlap or cross incoherently.
• If any diagram overlaps after render inspection, simplify, split, or remove the diagram; do not deliver an overlapped figure.