Higher-Dimensional Symbolic Dynamics: A Textile Framework For 3-graphs
arXiv:2607.29233
2026
Architecture
1 ideas extracted · analyzed Aug 31, 2026
What the math gives to ML
The paper offers a constructive local-to-global consistency mechanism: a three-dimensional textile system produces a 3-graph when colored maps satisfy pullback-square conditions and unique path-lifting properties. These conditions guarantee that differently ordered compositions of colored edges have unique equivalent factorizations, so local interchange relations extend coherently to global paths. A transferable neural-network design is a three-axis recurrent or state-space block trained to commute across axes, with explicit penalties for pairwise square defects and triple associativity defects; the main falsifiable benefit is reduced scan-order dependence and bounded long-horizon rollout disagreement.
Ideas from this paper
Unverified
2026
Use three learned state-transition operators corresponding to three data axes, and train them to satisfy the paper's pullback-style interchange rule. For every local pair of axes, two successive updates should reach the same square state; for triples of axes, all six update orders should agree. This reduces sensitivity to scan direction and limits long-horizon drift caused by inconsistent local transitions.
Useful6/10
Difficulty6/10
Novelty7/10