Spoke-and-Wheel teaching architecture: three quantum-specialist agents in parallel coordinated by a Synthesizer¶
Design · All designs
Evidence · 1 claim (1 for) · 1 study (1 design),q2· 0 of 1 report an effect size · 1 claim rests on one study
Description¶
ITAS's teaching architecture decomposes tutoring by task: three specialist agents (Video, Guidance, Code) analyze each student question in parallel, each reasoning only within its designated scope, and a Synthesizer Agent integrates their outputs using a priority hierarchy with code errors first, then conceptual gaps, then video references. The design responds to the prototype's task-boundary hallucinations, which the authors attribute to cognitive overload of a single overloaded agent. Specialization is said to reduce hallucination via cognitive load reduction, task interference elimination, and synthesis filtering. The three-agent count "emerged empirically; two agents conflated code and concepts, while four added coordination overhead without reliability improvement."
Design Implications¶
Context¶
Requirements¶
- Each agent must carry quantum-specific domain knowledge, e.g. a timestamp-indexed concept map for the Video Agent and a hand-curated Qiskit error catalog for the Code Agent
Constraints¶
- A controlled, blinded comparison against the prototype on the same input set remains future work
Target Learners¶
- graduate quantum information science students
Learning Goals¶
- conceptual understanding of quantum information science
- Qiskit implementation skills
- video lecture comprehension
Claims¶
Related Designs¶
Examples¶
-
Key Sources¶
- Iizalaarab Elhaimeur and Nikos Chrisochoides. (2026). From Prototype to Classroom: An Intelligent Tutoring System for Quantum Education. https://arxiv.org/abs/2604.24807
Page metadata
| Field | Value |
|---|---|
type |
design |
id |
spoke-and-wheel-teaching-architecture-itas |
title |
Spoke-and-Wheel teaching architecture: three quantum-specialist agents in parallel coordinated by a Synthesizer |
description |
ITAS's teaching architecture decomposes tutoring by task: three specialist agents (Video, Guidance, Code) analyze each student question in parallel, each reasoning only within its designated scope, and a Synthesizer A... |
status |
review |
generated |
by process:settle-candidates on 2026-10-11 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — iizalaarab-elhaimeur-and-nikos-chrisochoides-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.