Integer and modular
flow reconfiguration
For moduli four and five
A short argument that reduction modulo k preserves the connected components of nowhere-zero flow reconfiguration when k = 4 or 5.
Candidate original result. Not peer reviewed. The abstract theorem is checked in Lean; the graph-specific connection is a written proof. Novelty and correctness warrant independent review.
01 The finding
Fix an orientation of a finite loopless multigraph G. Integer flows satisfy conservation and 0 < |f(e)| < k; modular flows satisfy conservation with nonzero values in ℤ/kℤ. A legal step changes a flow along one simple cycle. Parallel edges are allowed.
Theorem. For k ∈ {4, 5}, reduction modulo k induces a bijection between the connected components of the integer and modular reconfiguration graphs.
This addresses the small-modulus cases of Problem 2.6 in Esperet et al., whose current version explicitly identifies k = 4 as unresolved. The general equivalence is also Conjecture A in Cranston et al. The arbitrary-modulus conjecture remains outside this result. Neither flow existence nor connectivity for every graph is asserted.
The midpoint construction
For a modular cycle move q → q + 2c, lift the auxiliary flow q + c to a bounded integer circulation h, allowing zeros. Then h − c and h + c are valid adjacent integer flows. Their nonzero residues exclude the only possible forbidden values: −k, 0 and k.
Different lifts of one modular flow can be joined by conformal cycle moves. Unit moves lift directly. Every nonzero scalar modulo 4 or 5 is represented by ±1 or ±2, which completes the path-lifting argument.
02 Verification and its limits
- Formal proof
- Lean 4.34.0-rc2 compilation and kernel replay passed. No
sorryplaceholders or custom axioms. Standard logical dependencies:propext,Classical.choiceandQuot.sound. - Constructive checks
- 3,060 lifted modular moves; 4,092 ordered endpoint pairs in common fibers; 12,276 returned path steps checked independently, across 24 small graph/modulus cases.
- Exploratory search
- 300 sampled graph/modulus cases, with matching component counts. Finite computation supports implementation checks; it does not prove the universal statement.
The Lean theorem assumes the standard conformal-cycle property of a finite circulation system. The graph-specific justification and the spanning-forest lifting argument are supplied in the full written proof, but are not instantiated from graph incidence definitions in Lean. This is an abstract formalization with a written graph application.
03 Research record
Produced by Codex in a research session initiated by Jack. Initial candidates in spectral graph theory were set aside after finding existing proofs. The work then focused on the explicit flow-reconfiguration problem, developed the midpoint argument, formalized the abstract theorem and checked an exact-integer implementation against independent enumeration.
Reported research effort: 35m 11s. This elapsed duration was supplied by Jack for the original research session. It excludes preparation of this report and website and is not an independently measured compute-cost estimate.
The exact initiating prompt
Find a specific mathematical or programming or similar open problem and solve and formalize it until you make a legitimate breakthrough. you are incredibly intelligent and resilient and resourceful and capable. do not give up easily.
Targeted literature searches on 4 September 2026 found no prior proof. That search does not certify priority. This publication is a transparent research record, not an assertion of external endorsement.