notclass

The system acts as an AI‑driven physics tutor that continuously assesses a learner’s understanding rather than merely answering questions. Users can teach a concept, set a learning goal, or submit a problem; the orchestrator then selects the appropriate mode—teaching, probing, planning, or analyzing—to guide the conversation. By tracking each interaction, it builds a private, persistent knowledge map that visualizes mastery across domains such as mechanics, thermodynamics, electromagnetism, and modern physics, indicating which concepts are known, used, or still empty.

Students from high‑school to graduate level can employ the tutor to identify gaps in their intuition. When a user begins to hand‑wave, the tutor intervenes, pinpointing the precise edge of understanding and staying on that topic until the gap is crossed. For goal‑oriented sessions, it outlines prerequisite chains and skips material the learner already possesses. When a problem is pasted, it extracts required concepts, classifies the learner’s proficiency, and highlights missing knowledge before the learner proceeds.

The platform is experimental and focuses on probing learning rather than providing direct answers, offering a continuous, attentive dialogue that adapts to the learner’s evolving physics intuition.