Mobius - AI Research Scientist

Job Description

Description Key Responsibilities : Design architectures for meta-learning, self-reflective agents, and recursive optimization loops. Build simulation frameworks for I behavior grounded in Bayesian dynamics, attractor theory, and teleo-dynamics. Develop systems that integrate graph rewriting, knowledge representation, and neurosymbolic reasoning. Conduct research on fractal intelligence structures, swarm-based agent coordination, and autopoietic systems. Advance Mobiuss knowledge graph with ontologies supporting logic, agency, and emergent semantics. Integrate I logic into distributed, policy-scoped decision graphs aligned with business and ethical constraints. Publish cutting-edge results and mentor contributors in reflective system design and emergent AI theory. Build scalable simulations of multi-agent, goal-directed, and adaptive ecosystems within the Mobius runtime. Required Qualifications Proven expertise in : Meta-learning, recursive architectures, and AI safety. Distributed systems, multi-agent environments, and decentralized coordination. Formal and theoretical foundations, including Bayesian modeling, graph theory, and logical inference. Strong implementation skills in Python (required), with additional proficiency in C++, functional or symbolic languages being a plus. Publication record in areas intersecting AI research, complexity science, and/or emergent systems. Preferred Qualifications Experience with : Neurosymbolic architectures and hybrid AI systems. Fractal modeling, attractor theory, and complex adaptive dynamics. Topos theory, category theory, and logic-based semantics. Knowledge ontologies, OWL/RDF, and semantic reasoners. Autopoiesis, teleo-dynamics, and biologically inspired system design. Swarm intelligence, self-organizing behavior, and emergent coordination. Distributed learning systems : Ray, Spark, MPI, or agent-based simulators. Technical Proficiency Programming Languages : Python (required), C++, Haskell, Lisp, or Prolog (preferred for symbolic reasoning. Frameworks : PyTorch, TensorFlow. Distributed Systems : Ray, Apache Spark, Dask, Kubernetes. Knowledge Technologies : Neo4j, RDF, OWL, SPARQL. Experiment Management : MLflow, Weights & Biases. GPU and HPC Systems : CUDA, NCCL, Slurm. Formal Modeling Tools : Z3, TLA+, Coq, Isabelle. Core Research Domains Recursive self-improvement and introspective AI. Graph theory, graph rewriting, and knowledge graphs. Neurosymbolic systems and ontological reasoning. Fractal intelligence and dynamic attractor-based learning. Bayesian reasoning under uncertainty and cognitive dynamics. Swarm intelligence and decentralized consensus modeling. Topos theory and abstract structure of logic spaces. Autopoietic, self-sustaining system architectures. Teleo-dynamics and goal-driven adaptation in complex systems. (ref:hirist.tech)