Noether Research
— · —
Noether Research

Noether Research

AI safety · Biosecurity · Biological intelligence · Complex systems

Abstract.

Noether Research is an independent research program focused on AI safety, AI-enabled biosecurity, and the mathematical structures connecting biological and artificial intelligence. Current work spans evaluation reliability and interaction safety in frontier AI systems, defensive biosecurity methods and screening frameworks, geometric methods applied to neural information propagation, and information-theoretic models of biological repair and reconstruction.

The program is led by a researcher with training in mathematical physics at MIT, astrophysics at the University of Cambridge, and computational molecular biology, with professional experience in life sciences strategy, decision analysis, and AI safety. Public summaries are intentionally limited where projects involve unpublished technical frameworks, sensitive evaluation methods, or dual-use considerations.

Research Program.
  1. 1.
    Frontier AI Safety and Scientific Reasoning
    Evaluation reliability, interaction safety, diagnostic frameworks, and responsible disclosure for advanced AI systems. Includes study of AI-assisted scientific reasoning, uncertainty tracking, and verification in AI-enabled research workflows.
  2. 2.
    AI-Enabled Biosecurity
    AI-assisted methods, defensive workflows, screening frameworks, institutional risk analysis, and near-term interventions for biosecurity.
  3. 3.
    Biological Intelligence and Information Fidelity
    Geometric methods applied to neural information propagation, information-theoretic models of biological repair and reconstruction, and comparative frameworks for non-human intelligence.
  4. 4.
    Complex Systems and Decision Dynamics
    Decision theory, crisis dynamics, infrastructure modeling, compute governance, and reliability in high-stakes complex systems under accelerating AI deployment.

Across these areas, the program focuses on reliability, information preservation, irreversibility, and the mathematical structure of high-stakes adaptive systems.

Restricted technical summaries are available to qualified reviewers, funders, and collaborators upon request. Some materials are withheld from the public site pending preprint release, responsible disclosure, or trusted review.