Open lab
What we release, the rules we work to, and the ways into the laboratory. Everything listed here is public and checkable; nothing is listed before it is.
Open by default
Every dataset listed here is deposited on Zenodo with a DOI and an open licence, so it can be cited and versioned rather than merely downloaded. Software is released under a free licence, and the two Moodle filters below are distributed through Moodle’s own plugin directory rather than only from a repository.
Released software
Both are Moodle filters published in the official Moodle Plugins Directory, licensed GPL 3.0 or later, and documented by a peer-reviewed paper. Being accepted into Moodle’s own directory means any institution can install them from inside its own site — the difference between a research prototype and a tool other people can actually run.
Analysis pipelines and replication packages for studies currently under
review are already public in the laboratory’s GitHub account, but they are not listed
individually here. A repository is added to this page when the study it supports is published,
so that every entry points at work a reader can actually read. Repositories awaiting that step
are tracked in data/repositories.json.
Rules we work to
Studies involving people are reviewed before data collection by the ethics commission of the institution that hosts them. For work at Kryvyi Rih State Pedagogical University, where most of the laboratory's members hold their appointments, that is the Research Ethics Commission of Kryvyi Rih State Pedagogical University — a standing collegial body chaired by the vice-rector for research and appointed by the rector for a three-year term. Its membership spans at least five faculties and includes a lawyer, a psychologist, a student representative and, where one agrees to serve, an external expert, so that a proposal is not reviewed only by people who share its assumptions. The commission works to the Law of Ukraine on Scientific and Scientific-Technical Activity, the Law on Higher Education, the Ethical Code of the Ukrainian Scientist and the Declaration of Helsinki.
An application states the aims and methodology, who the participants are — separately identifying minors, students, teachers, parents and people with special educational needs — the physical, psychological, social and economic risks and how each is mitigated, the form of informed consent, the confidentiality measures, how long data are kept and who may access them, and any conflict of interest. Reviewers are assigned on registration, the initial review takes ten working days, and a reasoned decision follows within thirty days: approved, approved with minor comments, returned for revision, or refused. Approval is a precondition for starting, and a substantive change to the protocol goes back to the commission.
The commission’s governing regulation is an internal university document and is not published online; it can be requested from the university.
The laboratory builds and evaluates generative models, and it also uses them in its own work. Both are governed by one principle: a generative model is an instrument, and the named authors carry complete responsibility for everything the work asserts.
The workshop series the laboratory organises apply the CEUR-WS policy on generative AI to every submission, and the Academy’s journals follow the COPE guidance on publication ethics.
Participants take part on informed consent, with parental or guardian consent for minors. Responses are coded on collection, and only de-identified data are deposited: no names, no contact details, no dates of birth, no identifiers that could re-identify a participant in combination. Where a dataset cannot be de-identified without destroying its research value it is not published, and the paper says so rather than quietly omitting the deposit.
The same rule governs this website. It carries professional information only — name, degree, academic title, position, institution and public research identifiers — and institutional contact addresses rather than private ones.
Datasets are released under CC BY 4.0 (one earlier deposit is CC BY-ND 4.0, marked as such in the list above); the Moodle filters are GPL 3.0 or later; the content of this site is CC BY 4.0. People are identified by ORCID and organisations by ROR so that a deposit can be attributed without relying on name matching.
The two projects funded from 2026 commit to preregistering their study protocols on OSF before data collection begins, and to depositing the resulting data openly. We state that as a commitment made in the funded proposals, not as a completed record: the laboratory has no preregistration to point to yet, and this page will link them when it does.
Join us
The laboratory has no separate staff list of its own: people work in it through their home institution, under an agreed annual plan. That makes several routes possible, and none of them requires a vacancy to be open.
Three doctoral researchers currently work in the laboratory, on topics inside its research programme — two of them are named executors on projects listed under Projects. Admission is through the doctoral programmes of the partner institutions; the research programme, the supervision arrangements and the venue come from the laboratory.
Master’s and undergraduate work at the Department of Computer Science and Applied Mathematics, on real components of funded projects — benchmark construction, evaluation harnesses, accessible interfaces, data pipelines — with the resulting artefacts released openly.
Researchers at other institutions join as affiliated members, keeping their own post while taking part in the laboratory’s programme, co-authoring and co-supervising. Several current members work this way.
We are looking for partners for Horizon Europe and Erasmus+ calls, and for national competitions with an international component. The laboratory already takes part in two Erasmus+ capacity-building projects and has held a competitive National Research Foundation grant.
A short message is enough to start. Tell us which of the six research themes your interest sits in, what you would want to work on, and include a CV with your ORCID. For doctoral enquiries, add a paragraph on the question you want to answer — not a full proposal; we would rather shape it with you.
We reply in English or Ukrainian. Enquiries from researchers displaced by the war are particularly welcome; several members work from outside their home city.
Doctoral admission and employment run through these: