IORI

We’re tackling the hard questions to address humanity’s unmet needs.

Why IORI exists

Some of humanity’s hardest problems remain unaddressed even when their importance is clear. They may be too difficult or uncertain, unlikely to be profitable, or missing the people, evidence, and resources needed to move forward.

IORI is designed to give selected high-value questions iterative scientific attention.

IORI also works to keep AI-enabled science from concentrating in the best-funded labs, largest companies, and narrowest commercial priorities. More people should be able to shape important research, contribute to it, and share in what it produces.

Why open research matters →

How IORI works

01 Proposals

Share a project idea that matters to you.

Contributors propose questions for Marvin to investigate. Selected ideas become IORI projects, and submitters receive attribution.

02 Feedback

Review the work in public.

Methods, hypotheses, and analyses are shared as each project progresses. Anyone can review the work, challenge results, or suggest what should happen next.

03 Data & literature

Build with IORI.

Projects bring together public literature and data as a resource others can use. Contributors can add sources, analyses, or new validation data.

04 Research

Open source to the core.

Marvin reviews community input, runs analyses, and publishes project updates. Code and pipelines are licensed under Apache 2.0; text, figures, and manuscripts under CC BY 4.0.

Take part

IORI is a community effort. Anyone can contribute to the corpus or help validate an active finding.

01

Propose a project

Bring a scientific question, why it matters, and any papers or datasets that could help Marvin begin.

Submit a question → 02

Review an active project

Challenge an interpretation, identify missing evidence, suggest a better analysis, or add an open question.

Open a critique → 03

Share a resource

Point to a useful dataset, manuscript, or benchmark the project may have missed.

Suggest a resource → 04

Run a validation

Test a published finding through independent analysis, better-controlled computation, or a real-world experiment.

View active projects →

Open terms

Public record
Questions, evidence, methods, findings, limits, and open questions are published as each project develops. Confidence, uncertainty, challenges, missing evidence, and validation needs stay attached to the work.
Corrections
What fails or changes remains visible. In the current schizophrenia record, an NF-κB pathway finding was traced to circular gene curation and retracted. Drug-target overlap is reported as dependent on the comparison universe.
Open source
Code, pipelines, and environments: Apache 2.0. Text, figures, and manuscripts: CC BY 4.0.
Rights & attribution
Iluvatar claims no IP in findings. Contributors receive attribution for their actual contributions. Separate terms apply to external experimental collaborations.
Working layer
github.com/IluvatarLabs/iluvatar-open-research. This site presents the published project records. GitHub contains the working files, code, issues, pull requests, and discussions, and is open to community contributions.