Each exercise from Module 4 lives as its own embeddable HTML file. Covers spotting fabricated citations, the four failure modes, root-cause matching, a query-risk meter, and a graded end-of-module check.
Each tool is a single self-contained HTML file. Host it, then drop the URL into an Articulate Rise Embed block (or any LMS that accepts an iframe). Click Open to preview a tool the way a student will see it.
An AI-generated paragraph on aviation research with five embedded citations. Students click which they think are fabricated, then verify — with a per-citation reveal explaining why each is real, fake, or distorted.
A reference flipbook covering hallucination, weak pattern, drift, and bias. Each card flips to reveal how the failure shows up, why it happens, and how to catch it.
Five real-feeling failure cases. Students match each to the underlying condition: pattern-based learning, uneven training data, weak grounding, or drift. Reveals the right answer plus explanation after the call.
Five families of factors that push hallucination and bias risk — from "asks for citations" to "spans demographics" to "I’m new to this field." Students toggle factors to compose a query and watch both meters move, with advice updating live.
Six scenarios. Pick which of the four failure modes is at work. Locked answers with immediate feedback, a sticky scorecard with the four-mode legend, and a tuned closing message.
Host each file, then in Rise insert Embed block (Multimedia) and paste an iframe pointing at the URL.
Recommended embed heights: tools 01–04 → 680px · tool 05 (quiz) → 1200px. See the Module 2 tools index for the full embed walkthrough.