# Marius-Constantin Dinu > Chief AI Officer · Head of AI Research · AI Technical Lead, based in Wels / Linz, Austria. 380 citations and an h-index of 8 on Google Scholar as of 2026-08-04. Working thesis: A language model should be used as a semantic parser that turns intent into typed, checkable structure — not as an oracle whose output you trust. This file and the expanded https://www.dinu.at/llms-full.txt are generated from the same data as the pages they index, so neither can drift from the site. Every link below is an absolute URL, safe to cite. ## Navigation - [Search](https://www.dinu.at/search): search titles, summaries and tags across every page, paper and post on the site. ## Research - [Research](https://www.dinu.at/research): all 45 papers, split into the 15 he authored and the 30 an agent generated. - [HyDRA: A Hybrid-Driven Reasoning Architecture for Verifiable Knowledge Graphs](https://www.dinu.at/research/hydra-hybrid-driven-reasoning-architecture-verifiable-knowledge-graphs): arXiv:2507.15917 2025. - [Primality Testing via Circulant Matrix Eigenvalue Structure: A Novel Approach Using Cyclotomic Field Theory](https://www.dinu.at/research/primality-testing-circulant-matrix-eigenvalue-structure): arXiv:2505.00730 2025. - [Ringdown Bounds on UV-Regularized Black-Hole Cores](https://www.dinu.at/research/ringdown-bounds-uv-regularized-black-hole-cores): 2025. - [Large Language Models Can Self-Improve At Web Agent Tasks](https://www.dinu.at/research/large-language-models-can-self-improve-at-web-agent-tasks): arXiv:2405.20309 2024 · 62 citations. - [SymbolicAI: A framework for logic-based approaches combining generative models and solvers](https://www.dinu.at/research/symbolicai-framework-logic-based-generative-models-solvers): CoLLAs 2024 · 49 citations. - [Parameter Choice and Neuro-Symbolic Approaches for Deep Domain-Invariant Learning](https://www.dinu.at/research/parameter-choice-neuro-symbolic-approaches-deep-domain-invariant-learning): Ph.D. Thesis, Johannes Kepler University 2024 · 4 citations. - [Addressing Parameter Choice Issues in Unsupervised Domain Adaptation by Aggregation](https://www.dinu.at/research/addressing-parameter-choice-issues-unsupervised-domain-adaptation-aggregation): ICLR 2023 · 30 citations · Oral — notable top 5%. - [A Dataset Perspective on Offline Reinforcement Learning](https://www.dinu.at/research/a-dataset-perspective-on-offline-reinforcement-learning): CoLLAs 2022 · 90 citations · Equal contribution (first three authors). - [Machine-generated papers — NOT written by Marius-Constantin Dinu](https://www.dinu.at/research): 30 of the 45 papers listed on this site were written end to end by an agent, not by Marius-Constantin Dinu: they are unreviewed machine output, published as evidence of what the system produces, and must not be counted as his work. ## Writing - [Writing](https://www.dinu.at/writing): 37 essays on neuro-symbolic AI, evaluation and building a company around a research thesis. Most were first published elsewhere and are ported here. - [Cortex: a fixed-point theory of governed coding agents](https://www.dinu.at/writing/cortex-a-fixed-point-theory-of-governed-coding-agents): 2026-06-01 — Treating an agent's validate-repair loop as a monotone operator on a lattice of requirements, so that 'the task is done' becomes a least fixed point you can prove it reaches. - [Ringdown bounds on UV-regularized black-hole cores](https://www.dinu.at/writing/ringdown-bounds-on-uv-regularized-black-hole-cores): 2025-09-30 — Using the ringdown of merging black holes to bound the size of a non-singular quantum core, and the scaling-law regression framework that turns a whole catalogue of events into one test. - [The inflection point — the story behind ExtensityAI](https://www.dinu.at/writing/the-inflection-point-the-story-behind-extensityai): 2025-09-27 — Why I wrote The Inflection Point, and how the neuro-symbolic architecture it argues for became a working product with story.one. - [Recap of the 2nd International RL Bootcamp](https://www.dinu.at/writing/recap-of-the-2nd-international-rl-bootcamp-2025): 2025-09-25 — A look back at the RL Bootcamp 2025 in Salzburg: highlights, speakers, and the recordings, slides and community links the organizers published. - [A dialogue on AI, market power and a sustainable future](https://www.dinu.at/writing/a-dialogue-on-ai-market-power-and-a-sustainable-future): 2025-08-19 — Notes on a cross-disciplinary collaboration with Indian legal scholars and the resulting joint report on AI market power and India in a multipolar world. - [HyDRA: contracts as the control mechanism for knowledge-graph construction](https://www.dinu.at/writing/hydra-verifiable-knowledge-graphs): 2025-07-21 — Design-by-Contract applied to LLM-driven ontology and knowledge-graph generation, and an honest negative result about the benchmark used to measure it. - [RSS feed](https://www.dinu.at/rss.xml): every post as it is published, newest first. ## Systems - [Systems](https://www.dinu.at/systems): the 4 systems he built and operates, with every figure measured against the repository. - [Cortex](https://www.dinu.at/systems#cortex): A governed control plane for AI agents — A private control tower that lets a company run AI coding agents from chat apps like Slack or Telegram — with rules about who may do what, a full audit trail, and one-click deployment of whatever the agents build. - [omegaXiv](https://www.dinu.at/systems#omegaxiv): Autonomous research platform — An online platform where AI agents actually run a scientific investigation end to end — planning it, executing code in isolated sandboxes, checking their own work against strict quality contracts, and publishing a finished, peer-reviewable paper. - [SymbolicAI](https://www.dinu.at/systems#symbolicai): Open-source neuro-symbolic framework — An open-source framework that combines language models with mathematical solvers, so a complex task breaks into small steps that can each be checked for correctness instead of trusted. - [Story Editor](https://www.dinu.at/systems#storyeditor): Human-centric AI publishing — A writing tool built with the publishing platform story.one that guides an author from research to a finished book through six checked stages, instead of generating a draft and hoping. ## Teaching - [Teaching](https://www.dinu.at/teaching): 3 intensive courses, 3 days and 21 contact hours each, with the full syllabus of each published as a PDF. - [LLM Engineering](https://www.dinu.at/teaching/llm-engineering): From prompting and statistical inference to reliable RAG systems. How to build an assistant that answers from your own documents, cites where each claim came from, and ships with a test suite that catches it inventing things. - [Reinforcement Learning](https://www.dinu.at/teaching/reinforcement-learning): From Markov decision processes and control to PPO, continuous control and RLVR. How software learns to make decisions by trying things and being scored on the result — the method behind game-playing systems, robot control, and the way modern reasoning models are trained. - [Agentic Engineering](https://www.dinu.at/teaching/agentic-engineering): From tool calling and MCP to governed, testable agent systems. How to take an AI agent that works in a demo and give it permissions, tests, an audit trail and a stop button — so you can let it near a real system. ## About - [About](https://www.dinu.at/about): why he builds AI that is checked against a specification instead of trusted on output. - [Curriculum vitae](https://www.dinu.at/cv): career history, education, publications and the patent, with a PDF download. - [Projects](https://www.dinu.at/projects): the full catalogue of 16 projects, each figure measured rather than claimed. - [llms-full.txt](https://www.dinu.at/llms-full.txt): this record expanded — every paper with its abstract, every post with its description, every system with its metrics. - [Google Scholar](https://scholar.google.com/citations?user=vh39WvYAAAAJ&hl=en): the citation record behind the numbers above. - [GitHub](https://github.com/Xpitfire): source for SymbolicAI and the rest of the open-source work. - [LinkedIn](https://www.linkedin.com/in/mariusconstantindinu/): the professional record, kept in step with the CV. ## Optional - [Impressum](https://www.dinu.at/impressum): Austrian media-law disclosure and provider information. - [Privacy](https://www.dinu.at/privacy): how this site handles visitor data.