Course materials
These are selected materials from my teaching and research. The discrete mathematics resources come from a classroom study that used two perturbed assignments with an AI literacy session between them. The files preserve the versions distributed or submitted at that time.
CSC 226: Discrete Mathematics
AI Literacy for Mathematical Reasoning
Lecture · PDF · 18 pages · Summer 2026
The lecture introduces ways to examine chatbot-generated mathematical reasoning. It presents three procedures: checking a proof, strengthening a prompt, and working through a problem with a chatbot.
Working with AI on Proofs
Exercises · PDF · 16 pages · Summer 2026
Exercises A–C give students practice with the three procedures. Labels appear in reference boxes, worked examples, and practice tasks so students can connect an action with its purpose.
Lab 3, Part 1
Assignment · PDF · 12 pages · Summer 2026
The first assignment in the classroom study includes perturbed counting, induction, and predicate logic questions. Students who used a chatbot were asked to submit the chat log with their work.
Final Exam Review Lab, Part 2
Assignment · PDF · 25 pages · Summer 2026
The second assignment follows the AI literacy session. Together, the two assignment parts also served as a final exam study guide.
CSC 484/584: Game AI
Applications of AI Literacy to GenAI-Assisted Game Development
Slides · PPTX · 55 slides · 17 September 2026
The guest lecture connects checking mathematical reasoning with reviewing an AI coding agent’s plans and output. A revised web reading edition provides explanations and notes alongside the lecture’s examples.
Study manuscript
Perturbed Assignments and Subgoal-Labeled AI Literacy Instruction in Undergraduate Discrete Mathematics
Submitted manuscript · PDF · 2 pages · 2026
This preprint describes the classroom study associated with the discrete mathematics materials. It was submitted to the SIGCSE TS 2027 Student Research Competition, and its chat-log analysis was still in progress when it was written.
Reading these materials
The original lecture uses simplified explanations of LLM training. For a more careful account, see the revised web lecture: next-token generation describes a decoding process, and post-training can change performance as well as presentation. Neither a confident explanation nor a successful small-case check establishes that a general proof is valid.
The procedures were taught as part of an ongoing study. Their presence in these materials does not establish that they improve learning. See the preprint for the study’s status and evidence.
Hallucination autopsy · Strong prompt structure · Collaborative prompting · Subgoal labels
For reuse, check the notice in each file. The manuscript PDF is a submission version; its conference template and placeholder DOI do not indicate acceptance or publication.