Larkspur Ridge High School Student newspaper One hour to press

The Second Edition

An interactive AI learning lab for grades 9–12

Reference Page

Sources and materials

What is real, what is invented, and where every cited claim comes from.

Sources cited in the lessons

The bracketed numbers beside claims in the lessons lead here. Every source below was opened and compared with the claim it supports on November 9, 2026. Links go to the publisher’s own page.

Machine learning

  1. [1]

    AI4K12 Initiative (jointly sponsored by AAAI and CSTA). “Five Big Ideas in AI”: Big Idea 3, Learning, and Big Idea 5, Societal Impact.

    https://ai4k12.org/

    Used here for: Framing for grades 9–12: computers can learn from data, and AI can affect society in positive and negative ways.

  2. [2]

    Google for Developers. “Decision trees.” Decision Forests course.

    https://developers.google.com/machine-learning/decision-forests/decision-trees

    Used here for: A decision tree is a model made of questions arranged as a tree; inner nodes hold conditions and leaves hold predictions.

  3. [3]

    scikit-learn developers. “1.10. Decision Trees.” scikit-learn User Guide (version 1.9.1 as viewed).

    https://scikit-learn.org/stable/modules/tree.html

    Used here for: Gini impurity is a common impurity measure used to score candidate splits when a tree is trained.

  4. [4]

    Google for Developers. “Datasets: Dividing the original dataset.” Machine Learning Crash Course.

    https://developers.google.com/machine-learning/crash-course/overfitting/dividing-datasets

    Used here for: A model should be tested on different examples from the ones used to train it, and duplicates of training examples should be removed from the test set.

  5. [5]

    Geirhos, R., Jacobsen, J.-H., Michaelis, C., Zemel, R., Brendel, W., Bethge, M., & Wichmann, F. A. (2020). “Shortcut Learning in Deep Neural Networks.” Perspective article, Nature Machine Intelligence. arXiv:2004.07780.

    https://arxiv.org/abs/2004.07780

    Used here for: The term “shortcut”: a decision rule that performs well on standard tests but fails to transfer to more challenging conditions.

  6. [6]

    Lapuschkin, S., Wäldchen, S., Binder, A., Montavon, G., Samek, W., & Müller, K.-R. (2019). “Unmasking Clever Hans Predictors and Assessing What Machines Really Learn.” Nature Communications. arXiv:1902.10178; doi:10.1038/s41467-019-08987-4.

    https://arxiv.org/abs/1902.10178

    Used here for: A published case of a real image classifier that leaned on a source tag printed on about one fifth of the horse pictures; removing the tag removed the “horse” answer.

Responsible use and journalism

  1. [7]

    National Institute of Standards and Technology (2024). Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST AI 600-1). doi:10.6028/NIST.AI.600-1.

    https://doi.org/10.6028/NIST.AI.600-1

    Used here for: Definition of confabulation: confidently stated but erroneous or false content, also called “hallucinations” or “fabrications.”

  2. [8]

    Society of Professional Journalists. “SPJ Code of Ethics.” The page viewed states it was revised October 4, 2026.

    https://www.spj.org/spj-code-of-ethics/

    Used here for: Journalists take responsibility for accuracy before releasing work, use heightened sensitivity with juveniles, and clearly label illustrations and reenactments.

  3. [9]

    Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. UNESCO.

    https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research

    Used here for: Global education guidance that calls for a human-centred approach to generative AI and for protecting data privacy.

Academic guidance

  1. [10]

    MLA Style Center. “How do I cite generative AI in MLA style?” Published 17 March 2023.

    https://style.mla.org/citing-generative-ai/

    Used here for: Cite a generative AI tool when you paraphrase, quote or incorporate content it created, and vet the secondary sources it cites.

  2. [11]

    MLA Style Center. “How do I cite generative AI in MLA style? (Updated and Revised).” Published 13 August 2025.

    https://style.mla.org/citing-generative-ai-updated-revised/

    Used here for: The updated post, which repeats that sources cited by a generative AI tool still need to be vetted.

Exercise materials (fictional)

These were written for this site so the lessons have something concrete to check. They are not real records, and the school policies among them are not general rules. The general principles they illustrate are the ones in the sources above.

MaterialWhere it appearsWhat it is
Photo records T01–T16, C01–C08, E01–E16, frame 214 and record N01Photo desk and front pageForty-three invented records with hand-entered yes/no values. No real photographs or people.
Larkspur Ridge Unified Board of Trustees, regular meeting record and start-time survey tableCopy deskAn invented district, meeting, vote and survey. The numbers were written so the arithmetic can be checked.
Prewritten “model responses” and the AI-assisted article draftCopy deskWritten by the site’s authors to show typical strengths and errors. Nothing is generated live and no AI service is called.
Biology notes and practice-question answer keyCopy deskA short invented study set. One answer-key entry is wrong on purpose.
The Second Edition newsroom standards, English 10 assignment policy and Chemistry lab-report policyStandards deskInvented school and class policies. They are not the rules of any real school; check your own teacher’s policy.
Regional robotics results sheet, story draft and picture optionsEditor’s challengeAn invented event, team and results. The “phone photo” and “AI image” are both drawings made for this site.

How the site is made

  • Illustrations. All pictures are original drawings generated by this site’s own code from simple shapes and dot patterns. No photographs and no stock or AI-generated images are used.
  • The photo sorter. A small decision-tree program that runs in your browser on the listed records. It is a teaching model and cannot read images.
  • “Model responses.” Prewritten by the site’s authors and labeled where they appear. The site calls no AI service.
  • Typefaces. Fraunces, Newsreader and Libre Franklin, each used under the SIL Open Font License 1.1 and served from this site.
  • Privacy. No accounts, no tracking and no personal data. Progress is stored only in your own browser; the Progress page explains how to erase it.