Experiment · L·01
All experimentsAES Lab
Experiments in automated essay scoring, comparing rubric-based heuristics with LLM-assisted grading on short-answer texts.
Objective
Explore whether LLM-based scoring can match consistent rubric grading while staying cheap enough for classroom use.
Status
In Development
Stack
Notes
Focused on short-answer and short-essay prompts; scoring output is always shown with a rationale for review.
Prototype. This experiment is not a production application.