Seonglae ChoIn research advising, having a "skill template" = having a good question framework
First, after you get the research proposal from someone, take a look and write down some ideas.
1. Problem Framing
- What is the core problem you’re trying to solve there?”
2. Contribution Check
- What would be new compared to existing work?
- If this works, who would care and why?
- What’s the paper-level contribution here?
3. Extension Path
- If you had 6 more months, how would you extend this?
- What’s the obvious next version of this?
- What breaks if you scale this up?
4. Technical Depth
- What is the main bottleneck in your current project?
- Where did things fail?
- Which assumption is most fragile?
5. Advisor Value
- What kind of help do you want from me?
- you need feedback on theory / experiments / positioning?
- Where are you stuck right now?
0. Talk
- What is your current project about?
- What is the main problem / bottleneck?
- How would you extend this beyond previous work?
- What’s the hardest part?
- Do you want help with?
Why & How approach
- Why does this work internally?
- Does this hold across models/domains?
- Does this still work at 10× scale?
- Where does this method fail?