Reproduce “EduGuard: A Safe RAG-Based LLM Tutor for Programming Education” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
S M Asif Hossain, Ruksat Khan Shayoni, M. F. Mridha, Jungpil Shin
EduGuard controls programming-tutor responses by checking course policy, selecting a teaching strategy, retrieving approved materials, and verifying claims against evidence. On the 600-query BILearn-CS benchmark, it achieved 90.1% correctness, 89.4% grounding, 90.8% rubric alignment, 4.9% hallucination, and 9.8% answer leakage. A 10-student pilot found higher immediate post-test accuracy, but does not establish long-term learning.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.