Maya-Morphe P3: Axon — Differentiable Bioelectric Edge Formation via Voltage-Guided Axonal Pathfinding in Morphogenetic Neural Networks

Abstract

Maya-Morphe P3: Axon extends the morphogenetic neural network framework developed in the earlier Maya-Morphe papers by asking a precise biological question: when a damaged cell recovers, does the specific pattern of connections it rebuilds affect tissue repair outcomes? To answer this, the authors introduce AxonNet, an architecture that adds a differentiable edge-formation module — using Gumbel-Softmax pathfinding — to guide recovering cells in forming new connections based on the surrounding bioelectric voltage field. Across 336 trials in Part 1 and 240 trials in Part 2, spanning cell populations from 29,056 to 266,071 cells, the study finds a scale-invariant revival threshold of approximately 0.162 and introduces Edge-Field Concordance (EFC) as the first quantitative metric for field-guided connectivity in morphogenetic computing, with EFC values ranging from −0.009 to −0.019 — indicating that recovering cells consistently probe slightly against the dominant voltage gradient in an exploratory manner. The threshold of ~0.162 is now confirmed across three independent experiments spanning all three Maya-Morphe papers, establishing it as an empirical substrate constant of this morphogenetic framework. This work is part of the Maya-Morphe Series from Nexus Learning Labs, Bengaluru, and advances the broader programme of using neuromorphic computation to model and simulate biological wound repair. Series: Part of the Maya-Morphe Series — morphogenetic neural networks for biological wound repair simulation. Bhaya Quiescence Law applies in modified form: dynamic-topology substrates show elevated β during active wound repair — biologically predicted and validated. Links: GitHub Repository (private — to request access: email research@nexuslearninglabs.in with subject Code Access Request — Maya-Morphe-P3 and your research context) | Interactive Dashboard | FAQ | Full Series Index — venky2099.github.io Nexus Learning Labs, Bengaluru · UDYAM-KR-02-0122422 · BHASKAR IN-0526-9452JSORCID: 0000-0002-3315-7907 · VAIRAGYA_DECAY_RATE = 0.002315 (embedded in all canonical hyperparameters)Canary: MayaNexusVS2026NLL_Bengaluru_Narasimha