work17

Pre-filter and crystallize research directions before running the full idea-pipeline. Narrows broad interests into 1-3 concrete, well-scoped ideas through constraint-driven exploration, direction deep-dive, and idea crystallization. Use when user says "filter ideas", "narrow directions", "选方向", "筛方向", "锁题", "帮我定个方向", "what should I work on", or wants to go from vague research interests to a focused, pipeline-ready idea.

Write Broader Impact sections for NeurIPS and other venues following Prof. Bingsheng He's (NUS) framework. Covers benefits, risks, uncertainties, and tractable/neglected/significant impacts. Use when writing impact or ethics sections.

Help find good research ideas and navigate early PhD years following Prof. Bingsheng He's (NUS) PhD Survival Guide. Covers the Year 1 checklist (10 activities), reading strategies, mentorship, research ideation, and building good habits.

Help prepare academic presentations, slides, and posters following Prof. Bingsheng He's (NUS) guidelines. Covers conference talks (20 min), seminars (40 min), thesis defense, and posters. Use when preparing any academic presentation.

Write a conference paper rebuttal following Prof. Bingsheng He's (NUS) proven rebuttal strategies. Analyzes reviews, categorizes concerns, drafts structured responses with evidence, and prepares confidential comments to SPC/AC. Use after receiving paper reviews.

Review a research paper draft as Prof. Bingsheng He (NUS) would. Applies his paper quality checklist, 5-figure rule, structure guidelines, and 28-point pre-submission checklist. Use when you want to review a paper before submission.

Guide writing of Qualifying Exam (QE) reports, Thesis Proposals (TP), and Final Theses following Prof. Bingsheng He's (NUS) guidelines and real exemplars. Covers structure, conversion from papers, plagiarism, and defense preparation.

Generate and rank research ideas given a broad direction. Brainstorms 8-12 ideas via external LLM, filters by feasibility, novelty, impact, and Prof. He's 4-dimension framework. Use when user says "找idea", "brainstorm ideas", "generate research ideas", "想点子", "what can we work on", or wants to explore a research area for publishable directions.

Full idea discovery pipeline: literature survey → idea generation → multi-dimensional screening → deep refinement. Use when user says "find ideas", "idea discovery", "full pipeline", "从零开始找方向", "找idea全流程", or wants the complete workflow from research direction to refined proposal.

Iterative deep refinement of a research idea via Problem Anchor + skeleton extraction + external LLM review. Turns a rough idea into a venue-ready proposal. Use when user says "refine idea", "打磨idea", "deepen this idea", "flesh out", "细化方案", or wants to turn a rough idea into a focused, concrete proposal.

Multi-dimensional screening of research ideas: novelty check + venue reviewer simulation + strategic fit assessment. Use when user says "screen ideas", "evaluate ideas", "review ideas", "novelty check", "查新", "筛选idea", or wants to rank and filter research ideas before committing to execution.

Search and analyze research papers across multiple sources to build a landscape map with identified gaps. Use when user says "literature survey", "文献调研", "find papers", "landscape map", "related work", "survey", or needs to understand the current state of a research area.

Use when the user wants to inspect a research idea from reference files, extract the core idea, build a detailed theory draft, and produce a lightweight experiment plan plus critical review notes in staged markdown outputs.

Use when given a specific Claude Code paper-idea folder containing `outputs/IDEA_DISCOVERY_REPORT.md`, `outputs/IDEAS_RAW.md`, `outputs/IDEAS_FILTERED.md`, and `refine-logs/FINAL_PROPOSAL.md`, and asked to produce a publication-grade `work.md` for the strongest idea, with idea audit, rigorous theory development, and resource-bounded experiment planning for a future ML paper.

Use when the user wants a first-principles search for low-resource AI research ideas with top-conference potential, covering foundation models, AI applications, and interdisciplinary directions under strict time and compute constraints.

Use when the user has one or more concrete research ideas and wants a strict reviewer-style evaluation before investing further, with emphasis on novelty, theory-practice linkage, implementation cost, story quality, and whether the idea tackles a genuinely hard problem.

Pre-filter and crystallize research directions before running the full idea-pipeline. Narrows broad interests into 1-3 concrete, well-scoped ideas through constraint-driven exploration, direction deep-dive, and idea crystallization. Use when user says "filter ideas", "narrow directions", "选方向", "筛方向", "锁题", "帮我定个方向", "what should I work on", or wants to go from vague research interests to a focused, pipeline-ready idea.

Use when the user wants to write a NeurIPS-style paper from existing idea, theory, and experiment materials, with emphasis on mature paper structure, detailed theory sections, and ready-to-replace figures.