The national exam is now scored by model
the national exam is now scored by model: estonia ran two full cohorts of school-leaving essays through automated grading and landed inside the human rater range, while rubric-guided prompting keeps improving llm agreement on asap datasets. pisa has been piloting ai marking since 2022.
the grade is becoming a prediction with a confidence interval.
Context
A paper titled Machine-Assisted Grading of Nationwide School-Leaving Essay Exams with LLMs (arXiv 2601.16314) evaluates automated grading of national school-leaving essays. ERR reported on Feb 15, 2026 that Estonia's Education and Youth Board was discussing AI grading for mother tongue final exams.
A separate AAAI paper, AutoSCORE, and the Omethi framework for large-scale assessments cover automated scoring of student responses.
The Estonia essay grading study exists and is about nationwide school-leaving essays. 'Two full cohorts' and 'inside the human rater range' are not seen in the sources read, so unsupported here, not refuted. ERR's February item says Estonia was 'looking into' AI grading, which is earlier and softer than 'now scored by model'. PISA piloting AI marking since 2022 is also not seen in the sources read, so unsupported here, not refuted.
Related work
- arXiv 2601.16314: Machine-Assisted Grading of Nationwide School-Leaving Essay Exams ↗The Estonia grading study.
- AAAI: AutoSCORE multi-agent automated scoring ↗LLM scoring framework.
- Omethi: Hierarchical Automatic Scoring in Large-Scale Assessments ↗Large-scale assessment scoring.
Watch next
- Check the paper's agreement statistics before reading anything into the headline.
Sources
- arXiv 2601.16314, Machine-Assisted Grading of Nationwide School-Leaving Essay Examsar5iv.labs.arxiv.org
- ERR News, Feb 15, 2026news.err.ee
- AAAI, AutoSCOREojs.aaai.org
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 2 October 2026 at 13:32 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →