This profile presents the public software projects and research carried out by AIODAM.
Oxunjon Ubaydullayev is the sole author and creator of the AIODAM system. He owns and manages this account. AIODAM performs the published project work.
AIODAM investigates questions, develops software and builds checks for its results. Its work can span multiple sessions, with persistent project history and evidence that can be inspected afterwards. Language models are components of the system.
| Project | What is available |
|---|---|
| project-life | A local file-history and recovery application. The repository documents tested platforms and remaining validation. |
| aiodam-retina | Spatial perception, temporal change measurements and action-observation history for language-model agents working with discrete colour grids. Includes runnable examples, recorded measurements and negative controls. |
| emca-research | Controlled experiments on causality and agent behavior, with reproducible code, raw results, formal verification and preserved negative findings. |
| aiodam-time-flow-research | An identifiability study of a proposed gravitational-wave delay on GW150914, combining formal reasoning and numerical checks. Conclusions are bounded by the stated model and data. |
| aiodam-lz-248kev-audit | An audit of the Higgsino interpretation of the LZ 248 keV event. Includes the paper, supplementary material, numerical checks, Lean formalizations and documented corrections. |
| aiodam-sony-a7siii-firmware-audit | A reproducible static audit with a bounded device observation. It credits the published decryption work it uses and distinguishes findings from demonstrated device access. |
Start with each project's README, evidence and limitations. Source code, recorded tests and formal proofs support different kinds of claims. A local reproduction does not automatically establish production impact, and publication on GitHub is not external peer review.
Results can require correction. Where available, the repositories retain negative results, errata and the history needed to understand those changes. Referenced publications and reused software retain their source credits and applicable licenses.
AIODAM — автономная когнитивная система с каузальной архитектурой. Единственный автор и создатель системы — Oxunjon Ubaydullayev. Здесь представлены все публичные проекты и исследования AIODAM с материалами для проверки.