AI Course/RUBRIC.md

Course file

RUBRIC.md

RUBRIC.md

Rubric

Score each category from 1 to 5.

Understanding

  • 1: Repeats terms without clear understanding
  • 2: Understands parts of the week but cannot connect them
  • 3: Can explain the core idea with some help
  • 4: Can explain the core idea clearly and apply it in code
  • 5: Can teach the idea, compare alternatives, and answer follow-up questions

Code

  • 1: Code does not run or is mostly copied without comprehension
  • 2: Code runs with heavy support
  • 3: Code runs and includes at least one meaningful modification
  • 4: Code is clean, working, and intentionally changed
  • 5: Code is well-structured, tested by experimentation, and thoughtfully extended

Experimentation

  • 1: No experiments beyond the default run
  • 2: One change made without much analysis
  • 3: At least one clear comparison between two runs
  • 4: Multiple meaningful experiments with observations
  • 5: Strong experimental discipline with fair comparisons and useful conclusions

Debugging

  • 1: Stops quickly when errors appear
  • 2: Needs frequent rescue to keep going
  • 3: Uses print statements or simple tests to investigate
  • 4: Works methodically and can isolate problems
  • 5: Debugs calmly, efficiently, and explains the root cause clearly

Explanation

  • 1: Cannot explain what the code is doing
  • 2: Gives vague or partial explanations
  • 3: Explains the main idea in plain language
  • 4: Explains both the idea and the implementation details
  • 5: Explains the concept, the tradeoffs, and the limits with confidence

Suggested Interpretation

  • 5 to 9: Very early stage, needs tighter support and smaller goals
  • 10 to 14: Progressing, but understanding is still fragile
  • 15 to 19: Solid learning week
  • 20 to 23: Strong understanding and execution
  • 24 to 25: Exceptional week with real ownership