Reproduce “TutorLoop: Regulating Student Learning Behaviors via Sensor-in-the-Loop Generative Feedback” 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
Songlin Xu, Xinyu Zhang
TutorLoop uses webcam-derived attention and workload estimates to guide feedback. A reinforcement-learning policy trained with simulated students selects among four actions over time, while an LLM refines the chosen message. In a randomized study of 187 participants, Full achieved mean semantic similarity of 0.284, exceeding DRL Only and TutorUp; evidence is limited to recall after one 10-minute task.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.