Search
Everything, in one place.
Nothing matches that. Try a broader word — a topic, a venue, or an author's surname.
Paper
- A Contradiction-Aware Survey Framework for Multi-Objective Decision SupportMulti-objective decision support now spans classical scalarization, Pareto-evolutionary search, outranking-based multicriteria ranking, and uncertainty-aware formulations, but prac
- AutoTW-ASP: Automatic Low-Treewidth Encoding Synthesis and Backend Routing for Neurosymbolic ASPNeurosymbolic answer set programming (ASP) systems often achieve exact semantics only after manual encoding redesign, which makes runtime gains difficult to scale beyond expert-cur
- AutoTW-ASP: Automatic Low-Treewidth Rewrite Synthesis and Uncertainty-Aware Backend Routing for Exact Neurosymbolic ASP TrainingRecent neurosymbolic Answer Set Programming (ASP) pipelines frequently rely on manual encoding redesign to exploit treewidth-sensitive exact inference backends, creating a reproduc
- Benchmarking and Selecting State-of-the-Art Modern Fourier Transformation MethodsFourier transforms remain critical to scientific simulation, signal analysis, and modern machine learning workloads, yet practical performance leadership is conditional on workload
- Calibrated Hybrid Evaluation of Quantum Reservoir Classification Under Finite-Shot and Simulability ConstraintsQuantum reservoir computing has shown repeated empirical promise for representation learning, but the evidence base for robust quantum advantage in image classification remains fra
- Certified Regime Mapping for Quantum Reservoir Computing Under Parity-Constrained EvaluationQuantum reservoir computing is frequently evaluated with point-estimate accuracy gains that confound representation effects, readout parity, and computational cost. We present a hy
- Compositional Security Control for AI-Assisted Coding Workflows: A Threat-Model-Grounded Security–Productivity FrontierAI-assisted coding systems now influence repository edits, dependency selection, and deployment decisions, which creates coupled attack surfaces spanning prompt-channel abuse, supp
- Conditional Constrained Routing and Metric Bridging for SymbolicAI Workflows Under CPU-Only BudgetsModular language-agent systems increasingly combine large language models, tool calls, and symbolic operators, but objective design and evaluation practice remain misaligned: traje
- Conservative Offline RL with Uncertainty-Aware Policy ImprovementWe study conservative offline reinforcement learning with uncertainty-aware policy improvement under a tight compute budget. The goal is to combine conservative value regularizatio
- Contract-Governed Multi-Agent Graph Orchestration for Long-Horizon Autonomous Research PipelinesLong-horizon research automation needs stronger semantics than prompt chaining or schema-only artifact passing. We study a restricted contract-governed formulation of Quarks in whi
- Cortex: A Fixed-Point Theory of Governed Coding AgentsA coding agent is a large language model (LLM) wrapped in a control loop — a harness — that lets it plan, write, and execute code. Raw harnesses optimize next-token capability, not
- Curiosity-Conditioned Goal-Optimal Reinforcement LearningGoal-conditioned reinforcement learning often faces a practical tension: intrinsic novelty bonuses accelerate discovery in sparse and deceptive environments, but poorly controlled
- Dependence-Aware Multi-Head Activation Monitoring for Distribution Shift and OOD ReliabilityModern neural systems frequently fail under deployment shift because confidence-only diagnostics underrepresent hidden changes in internal activations, and static benchmark metrics
- Detecting Physical and Procedural Bias in Lottery Draws: A Number-Theoretic and Statistical StudyPhysical lottery systems are designed to approximate uniform sampling without replacement, yet practical implementations involve latent mechanical and procedural factors that can i
- Do Sunspot Cycles Causally Affect Social Tensions and Population Harm in Developing Countries?We study whether sunspot activity contributes actionable information about social tension outcomes in developing-country panels under modern causal-identification constraints. Buil
- Dual-Timescale Task-Agnostic Activations for Continual Learning: Stability Guarantees and Boundary-Case EvidenceContinual learning systems are increasingly deployed in settings where data distributions evolve while labels, environments, and downstream requirements remain nonstationary. In th
- Durable Engraftment Modeling for Stem-Cell-Derived Islet Replacement in Type 1 DiabetesType 1 diabetes (T1D) remains a paradigmatic autoimmune disease in which loss of beta-cell function causes dysglycemia, severe hypoglycemia, and lifelong dependence on exogenous in
- Entropy-Aware Memory Systems for Continual Learning: Balancing Neuroplasticity and Stability Under Stochastic WorkloadsContinual learning systems are increasingly limited by memory behavior rather than arithmetic throughput: the same memory substrate must support stable recall and adaptive updates
- Glucose-Responsive Insulin Design via Hybrid Machine Learning, Molecular Dynamics, and Pareto SelectionType 1 diabetes management remains constrained by insulin therapies that are dosed externally and therefore cannot adapt in real time to changing glycemic states. This gap motivate
- Inner-Shell Raman X-Gate Tradeoffs for a Neutral ¹⁷¹Yb Nuclear Qubit at Fixed Optical PowerWe study single-qubit X-gate feasibility for a 171 Yb nuclear-spin qubit encoded in the 3 P0 (F = 1/2) manifold and driven by Raman coupling through an inner-shell-excited J = 2 in
- Interference-Gated Dynamic Activation for Task-Agnostic Continual Learning: A Formal-Empirical Audit of Stability, Forgetting, and Failure RegimesContinual learning methods frequently reduce forgetting by adding replay, regularization, or routing constraints, yet activation functions are usually treated as fixed nonlineariti
- Local-Energy Embedding for Critical Control in 3D Navier–Stokes: A Proof Program and Quantitative CriteriaWe develop a proof program that targets a quantitative bridge between scale-invariant local energy control and global critical L∞ 3 t Lx bounds for the three-dimensional incompress
- Material Signatures for Antineutrino-Based Detectability of Covert Fissile Production in Fusion ReactorsAntineutrino monitoring is a promising route for early safeguards signals, but fusion-adjacent deployment requires robustness to prior disagreement, detector nuisance variability,
- Navier–Stokes Regularity via Critical Norm TrackingWe study 3D incompressible Navier–Stokes flow on a periodic box and design a diagnostic suite that links classical Prodi–Serrin mixed norms, weak-Lp (Lorentz) proxies, scaling-inva
- Noise-Biased Surface Code Thresholds Under Realistic Gate SetsBiased-noise threshold claims for surface-code families are often expressed in terms of the nominal hardware dephasing ratio η, yet realistic gate decompositions, measurement asymm
- Parity-Constrained Quantum Reservoir Computing for Image Classification: Formal Guarantees and Staged Simulation EvidenceQuantum reservoir computing has recently reported encouraging image-classification performance, yet many claims remain sensitive to fairness controls, measurement-policy confounds,
- Parity-Locked Quantum Reservoir Computing for PCA-Encoded Image Classification: Robust Advantage, Entanglement Frontiers, and Operator–Dynamics AttributionQuantum reservoir computing for image classification is currently constrained by a reproducibility problem: many reported improvements can be explained by uneven preprocessing, rea
- Quantum Reservoir Computing Under Comparator Parity: Regime-Conditioned Advantage, Entanglement Effects, and Kernel-Null BoundariesQuantum reservoir computing (QRC) is often evaluated with heterogeneous comparator strength, making it difficult to determine whether reported gains are genuinely quantum-mechanist
- Representation-Conditioned Synthesis for Multi-Objective Decision Support: Formal Comparability, Evidence Grading, and Boundary DiagnosticsMulti-objective decision support is fragmented by a representation mismatch: Pareto-set methods, outranking relations, and scalarization-based methods emit different mathematical o
- Simulability-Aware Quantum Reservoir Computing for Image Classification under Matched Readout FairnessQuantum reservoir computing for vision tasks is often discussed in terms of empirical gains without an equally explicit separation between predictive improvement and computational-
- Software Commoditization in the AI-First Economy: From Implementation Moats to Contribution MarketsAI-assisted software generation changes the economics of software scarcity. We argue that value does not disappear when recognizable interfaces and workflows become easier to repro
- Stability-Aware Bilevel Source Dataset Selection for Importance-Weighted Least Squares in Unsupervised Domain AdaptationImportance-weighted least squares (IWLS) is widely used to correct covariate shift in unsupervised domain adaptation, yet most prior work assumes that an appropriate source dataset
- HyDRA: A Hybrid-Driven Reasoning Architecture for Verifiable Knowledge GraphsThe synergy between symbolic knowledge, often represented by Knowledge Graphs (KGs), and the generative capabilities of neural networks is central to advancing neurosymbolic AI. A
- Primality Testing via Circulant Matrix Eigenvalue Structure: A Novel Approach Using Cyclotomic Field TheoryThis paper presents a novel primality test based on the eigenvalue structure of circulant matrices constructed from roots of unity. We prove that an integer n > 2 is prime if and o
- Ringdown Bounds on UV-Regularized Black-Hole CoresSpacetime singularities in black-hole solutions signal a breakdown of the classical description at high curvature. We analyze a minimalist UV-regularized black-hole model with a si
- Large Language Models Can Self-Improve At Web Agent TasksTraining models to act as agents that can effectively navigate and perform actions in a complex environment, such as a web browser, has typically been challenging due to lack of tr
- SymbolicAI: A framework for logic-based approaches combining generative models and solversWe introduce SymbolicAI, a versatile and modular framework employing a logic-based approach to concept learning and flow management in generative processes. SymbolicAI enables the
- Parameter Choice and Neuro-Symbolic Approaches for Deep Domain-Invariant LearningAs artificial intelligence (AI) systems advance, we move towards broad AI: systems capable of performing well on diverse tasks, understanding context, and adapting rapidly to new s
- Addressing Parameter Choice Issues in Unsupervised Domain Adaptation by AggregationWe study the problem of choosing algorithm hyper-parameters in unsupervised domain adaptation, i.e., with labeled data in a source domain and unlabeled data in a target domain, dra
- A Dataset Perspective on Offline Reinforcement LearningThe application of Reinforcement Learning (RL) in real world environments can be expensive or risky due to sub-optimal policies during training. In Offline RL, this problem is avoi
- Align-RUDDER: Learning From Few Demonstrations by Reward RedistributionReinforcement learning algorithms require many samples when solving complex hierarchical tasks with sparse and delayed rewards. For such complex tasks, the recently proposed RUDDER
- Reactive Exploration to Cope with Non-Stationarity in Lifelong Reinforcement LearningIn lifelong learning, an agent learns throughout its entire life without resets, in a constantly changing environment, as we humans do. Consequently, lifelong learning comes with a
- InfODist: Online distillation with Informative rewards improves generalization in Curriculum LearningCurriculum learning (CL) is an essential part of human learning, just as reinforcement learning (RL) is. However, CL agents that are trained using RL with neural networks produce l
- The balancing principle for parameter choice in distance-regularized domain adaptationWe address the unsolved algorithm design problem of choosing a justified regularization parameter in unsupervised domain adaptation. This problem is intriguing as no labels are ava
- XAI and Strategy Extraction via Reward RedistributionIn reinforcement learning, an agent interacts with an environment from which it receives rewards, that are then used to learn a task. However, it is often unclear what strategies o
- Artificial Intelligence, Market Power and India in a Multipolar WorldThe artificial intelligence landscape isn't just about technological advancement - it's fundamentally about power concentration and market control. Our latest comprehensive researc
Essay
- Agentic AI for Scientific Discovery — the RL Bootcamp 2026 slidesThe 3rd International RL Bootcamp runs 16–18 September 2026 in Salzburg, free to attend. Here is the deck for my Friday talk, embedded to click through in full, plus what it argues and where the recordings land.
- The moat was never the softwareIf a product can be specified precisely enough to regenerate and check, an AI-assisted team can now rebuild it in a week. That is a narrower claim than it sounds — and it points at where the value actually sits.
- Cortex: a fixed-point theory of governed coding agentsTreating 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 coresUsing 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 ExtensityAIWhy 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 BootcampA 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 futureNotes 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 constructionDesign-by-Contract applied to LLM-driven ontology and knowledge-graph generation, and an honest negative result about the benchmark used to measure it.
- Entering the three-body problemInspired by sci-fi's The Three-Body Problem, this post explores chaos, system dynamics, and how AI research automation can master complexity.
- Primality testing via circulant matrix eigenvalue structureA prime is exactly an integer whose circulant matrix has a minimal polynomial with two irreducible factors over the rationals. A clean characterisation, and a slow algorithm.
- Empowering research innovationHow the neuro-symbolic stack behind SymbolicAI became a research automation platform, and what happened when we pointed it at an open question in primality testing.
- The benchmark illusionAI models ace benchmarks yet fail real-world tests — a critical assessment in evaluations and why we need better frameworks to measure intelligence.
- PyFlow.ts: bridging ML research with productionHow PyFlow.ts exposes Python ML code to TypeScript with one decorator, removing the API, typing and client boilerplate of the last-mile deployment step.
- Year one in reviewRecap of ExtensityAI's first year: origins as a neuro-symbolic research framework, the team, milestones, publications, and plans for 2025.
- Parameter choice and neuro-symbolic approaches for deep domain-invariant learningMy doctoral thesis at JKU Linz: how to pick hyperparameters for domain adaptation when the target has no labels, and what to do when you cannot train at all.
- Large language models can self-improve at web agent tasksHarvesting an agent's own WebArena trajectories, filtering them without labels, and fine-tuning on the result — plus two metrics for what a benchmark score hides.
- SymbolicAI: industry recognition and community impactRecognition that the SymbolicAI framework and its research paper received from AI researchers, industry figures, media outlets, and the community.
- Announcing the SymbolicAI paperThe research paper behind SymbolicAI: a modular framework for logic-based approaches that compose generative models with solvers, plus a benchmark for evaluating LLMs in AI-centric workflows.
- A neuro-symbolic perspective on large language modelsThe Symbolic API: using LLMs as the core of a neuro-symbolic stack, decomposing hard tasks into simple zero-shot operations and recombining them.
- InfODist: online distillation with informative rewardsWhy curriculum-learning agents generalise badly to the next task, and why the culprit is the states they explore rather than the non-stationarity they endure.
- Addressing parameter choice issues in unsupervised domain adaptation by aggregationExtending weighted least squares to vector-valued functions, with a target error asymptotically no worse than twice the unknown optimal aggregation.
- A dataset perspective on offline reinforcement learningTwo measures — SACo for exploration, TQ for exploitation — that characterise a behavioural policy's dataset and predict which offline RL algorithms will work on it.
- Reactive exploration to cope with non-stationarity in lifelong reinforcement learningTracking continual domain shifts as they happen: why policy-gradient methods adapt faster than Q-learning when the environment keeps moving.
- Align-RUDDER: learning from few demonstrations by reward redistributionReplacing RUDDER's LSTM with a profile model from multiple sequence alignment, so reward redistribution works from a handful of demonstrations.
- The balancing principle for parameter choice in distance-regularized domain adaptationBorrowing the balancing principle from ill-posed inverse problems to justify the regularization parameter when the target domain has no labels.
- Lighter: dependency injection for PyTorchPython borrowed little from the aspect-oriented practices that matured in Java and C#. Lighter brings dependency injection to PyTorch projects.
- JKU AI overview videoA short overview of what AI makes possible, produced at Johannes Kepler University Linz.
- XAI and strategy extraction via reward redistributionUsing Align-RUDDER as an interpretability method: multiple sequence alignment surfaces the key events an agent relies on, and those events are usually human-readable sub-tasks.
- Align-RUDDER: reward redistribution from few demonstrationsThe first write-up of Align-RUDDER: two modifications to RUDDER that let profile models built from as few as two demonstrations carry the reward redistribution.
- Overcoming catastrophic forgetting with context-dependent activationsMaster's thesis: context-based gating that switches pathways through a network, cutting destructive interference on long task sequences.
- Imagine KaraA thank-you to the team behind Imagine Kara as the project wound down.
- Seamlessly entering the crypto world with ApollonA short pointer to a longer write-up on the Apollon project and getting started with masternodes.
- Earning almost one Bitcoin with masternodes in three weeksHow hosting Apollon masternodes worked out over three weeks in 2018, and how the project's search interest compared to the top ten coins.
- Uni Swift projectA university iOS app for managing photos, built around actually finding the right image again.
- Deep Learning ScriptA DSL for Caffe that cuts the line count of prototyping a network architecture, with a VS Code extension that transpiles to Caffe Script.
- Deep learning: where to start and how to dig deeperTwo reference notes from 2016 — an overview of the field, and a practical guide to running deep learning on Ubuntu with Docker.
- Operation PhrikeA VR combat-simulation study: Oculus Rift plus Unreal Engine, with sensors reading the test subject's stress level.
- Internship report: handwritten character recognitionA Xamarin app that classifies handwritten characters with a neural network, compared against a support vector machine. Written at Siemens Corporate Technology.
- Bachelor thesis: cross-language integration for the CLRCompiling JavaScript to .NET IL by pairing a Coco/R-generated parser with Roslyn's syntax trees.
System
- CortexA 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.
- omegaXivAn 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.
- SymbolicAIAn 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 EditorA 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.
Project
- CortexA 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.
- omegaXivAn 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.
- EngramThe part of the platform that remembers: every conversation, coding session and dropped file becomes a searchable, linked note, and work the team keeps repeating is turned into a reusable recipe overnight.
- GauntletA test rig that answers a question nobody else measures: how much of an AI coding agent's safety and output quality comes from the model, and how much from the scaffolding wrapped around it.
- DentateTakes the benchmark's score as a reward signal and trains a small model that runs on one machine to do the work a frontier model was doing.
- SymbolicAIAn 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.
- SemaA new programming language in which asking a model a question and proving a property about the answer are both ordinary, first-class instructions — so the checks cannot be skipped.
- SynapseTurns a paragraph of instructions into a checklist the agent cannot quietly drop items from, then rules on whether the finished work actually satisfies it.
- NovaA development environment that runs in a browser tab, where people and AI agents work on the same project side by side and every change is reviewed before it ships.
- Cortex VantageA voice-first app an executive opens from a chat message to ask a question about their own company and get an answer that shows its sources, states what is uncertain, and proposes a next step.
- Graviton TheoryA music project — a ten-track album called Ghost in the Frame — with its own site, where every song goes up with its lyrics and its artwork the moment it is done instead of waiting for a release date.
- MyflixA private, streaming-style app for a personal film and series collection, on the TV as well as in the browser — built to find out how much of a familiar product an AI-assisted team can actually produce, and which parts turn out not to be the valuable ones.
- Story EditorA 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.
- Sema Epistemic Flight LabA flight simulator where every number on screen can be traced back to the model and the assumption that produced it, instead of just looking plausible.
- Vehicle Engineering SimulationA prototype showing the same idea applied to car engineering: airflow around a vehicle and its assembly sequence, with every derived figure labelled as an assumption rather than a measurement.
- Biological SimulationA prototype that models biology at three zoom levels — molecules, cells, tissue — as a single program, so a result at one scale stays linked to what produced it at another.
- Math Riddle BookA small interactive prototype: a book of maths riddles you actually play in the browser rather than read.
Course
- LLM EngineeringHow 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 LearningHow 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 EngineeringHow 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.
Talk
- Reinforcement Learning — Where We Are and What's NextInternational RL Bootcamp, 2025.
- KI-Startup-Founder über OpenAI, DeepSeek und Artificial General Intelligencebrutkasten, 2025.
- Extensity AI — The tool that automates research workstartup.ro, 2025.
- Building reliable and explainable AI agent systemsWhy SymbolicAI treats a language model as a semantic parser inside a workflow rather than the workflow itself, and how you measure whether an individual reasoning step was actually correct.
Page
- HomeStart here.
- AboutBackground, roles and how I work.
- ResearchPublications, preprints, the patent and machine-generated papers.
- SystemsCortex, omegaXiv, SymbolicAI and the Story Editor, with measured figures.
- ProjectsEverything built, with status and stack.
- ConsultingAdvisory, technical diagnostics, project delivery and fractional Chief AI Officer work, with published rates.
- TeachingThree-day course syllabi and university teaching.
- WritingEssays on neuro-symbolic AI, reinforcement learning and building systems.
- CVFull curriculum vitae, downloadable as a PDF.
- MusicGraviton Theory — two cyber nu-metal records, with a reel from each.
- ImpressumLegal notice and contact details.
- PrivacyWhat this site measures, and how to withdraw consent.