Navigating Europe’s Artificial Intelligence Act: Application of LLMs in classrooms
Bibliographic record
Abstract
In 2018, OpenAl introduced the first version of the Generative Pre-trained Transformer (GPT), revolutionizing the future of Large Language Models (LLMs). This model demonstrated the potential of pretraining large-scale models with vast text data and then fine-tuning them for specific tasks to the public. LLMs have quickly penetrated educational environments, aiding students from various disciplines in tasks ranging from initiating research to drafting essays. While the latter may breach academic integrity, the former is highly beneficial, especially for exploring new areas or ideas. Comparatively, the user interactions with GPT might resembles initially with that of search engines, despite technological differences, as both provide answers to queries, often reflecting archived as well as mainstream views. The historical evolution of search engines, from Archie's database matching to Google's relevance-based ranking, highlights similar ethical considerations faced by both technologies. The development of search engines underscored the importance of accessible information, a principle equally relevant to GPT and LLMs today. This paper is written in light of recent coming into force of European Union's legal framework on artificial intelligence for the purpose of examining adoption of LLMs in classrooms, and argues for balanced regulations across jurisdictions that acknowledge both the immense educational potential of LLMs and the need for adherence to legal and ethical standards.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.011 | 0.038 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.004 | 0.009 |
| Scholarly communication | 0.008 | 0.011 |
| Open science | 0.002 | 0.007 |
| Research integrity | 0.004 | 0.004 |
| Insufficient payload (model declined to judge) | 0.009 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".