Minimal Axiom Sets for Kerr Geodesic Conservation Laws: A Machine-Verified Audit Using Aristotle
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Abstract We present, to our knowledge, the first systematic machine-verified audit of foundational assumptions used in classical proofs of Kerr geodesic conservation laws. The method is intentionally adversarial: for each target theorem, we submit an under-specified Lean 4 theorem stub to Aristotle (Harmonic’s AI theorem prover), then record which hypotheses, definitions, or infrastructural objects Aristotle must inject to close the proof attempt. These injected artifacts are treated as explicit witnesses of assumptions that are usually implicit in coordinate-free presentations. We formalized all four standard conservation laws in Boyer–Lindquist coordinates (rest mass, energy, axial angular momentum, and Carter constant). The central result is a formal partition into two classes. Energy and angular momentum, both tied to Killing vectors, are derivable from a common axiom set and yield isomorphic proof certificates. By contrast , the Carter constant, tied to a rank-2 Killing tensor, requires an irreducibly new axiom: across seven submissions and two independent tensor formulations, Aristotle either introduces a circular conservation premise or collapses derivative infrastructure to trivi-alize the target equation. Secondary results include three unstated C 2 regularity assumptions in energy conservation, a syntax-forced clarification of the symmetrized Killing-tensor equation, and diagnosis of Aristotle’s intermediate “Form B” tensor as mathematically incorrect by componentwise verification. Artifacts are available at github.com/abdulrahimiqbal/KerrFormalization.
Fields of Study & Subjects
Relativity and Gravitational Theory
Logic
programming
and type systems
Advanced Differential Geometry Research
Publication details
- Journal / source
- Research Square
- Publication date
- 26 March 2026
- Language
- EN
- Document type
- Preprint