Rubric & Feedback Config
Configure how the AI grades submissions and generates feedback for students.
The rubric settings control how the AI evaluates student submissions. Access these settings from the Rubric section in your module.
Tip: Take time to configure your rubric before students start submitting. This ensures consistent grading from the start.
Choose how the AI communicates with students:
Encouraging
Positive, supportive feedback that motivates students
Neutral
Objective, balanced feedback focusing on facts
Critical
Direct feedback highlighting areas for improvement
RAG (Retrieval-Augmented Generation) settings control how the AI uses your documents:
- Context retrieval - How much document context to include when grading
- Relevance threshold - Minimum relevance score for retrieved content
- Source citations - Whether to include document references in feedback
Add custom instructions to guide the AI grading:
- Specific criteria you want the AI to evaluate
- Keywords or concepts that must be included
- Common mistakes to watch for
- Grading scale preferences
Example Custom Instruction:
"Focus on whether the student demonstrates understanding of the core concepts. Deduct points for factual errors but be lenient on formatting. Encourage students to provide examples from the course materials."
Configure your grading scale:
- Points per question - Maximum points for each question
- Passing threshold - Minimum score to pass
- Grade boundaries - A, B, C, D, F thresholds
The AI will use this scale when assigning scores. You can always manually adjust grades after review.