MEDS dataset of 28,000 LLM runs across four mathematics-education tasks¶
Product or Programme · All products and programmes
Evidence · 2 claims (1 for, 1 mixed) · 2 studies (2 associational),q2· 0 of 2 report an effect size · 2 claims rest on one study
Description¶
A publicly released dataset by Esposito et al. containing 28,000 structured multimodal LLM runs across mathematics interviews, psychometric questionnaires, cognitive networks, and quiz items for studying AI mathematics reasoning and learner-related constructs.
Components¶
Claims¶
- LLM difficulty predictions for fraction items are less stable across repeated runs than for the overall item set [~W]
- LLM digital shadows reproduce the human pattern of positive science perceptions anchored in negative mathematics associations [+W]
Related Products and Programmes¶
- Large-scale field study dataset and code release for AI-generated exam evaluation
- TURTLEAI benchmark
- 398-item evaluation dataset of teacher-requested geometric diagrams from Coteach
- The Pedagogy Benchmark
Key Sources¶
- Esposito, N., Tricarico, A., Porzio, L., Aghazadeh Ardebili, A., & Stella, M. (2026). Math Education Digital Shadows for Investigating Learning with GenAI: Mathematics Performance, Anxiety, and Confidence in LLMs. arXiv preprint. https://arxiv.org/abs/2604.27618. doi:10.48550/arxiv.2604.27618
Page metadata
| Field | Value |
|---|---|
type |
product |
id |
meds-dataset-of-28000-llm-runs-across-four-mathematics-education-tasks |
title |
MEDS dataset of 28,000 LLM runs across four mathematics-education tasks |
description |
A publicly released dataset by Esposito et al. |
status |
review |
product_kind |
dataset |
generated |
by process:settle-candidates on 2026-10-11 |
trust |
unverified — no one has confirmed this page's content |
Frontmatter is the page's machine-readable record — see the Schema & Guide.