Eymen Ipek

Eymen Ipek

Odrzywolek (2026) recently introduced the Exp-Minus-Log (EML) operator eml (x, y) = exp(x) - ln(y) and proved constructively that, paired with the constant 1, it generates the entire scientific-calculator basis of elementary functions; in this sense EML is to continuous mathematics what NAND is to Boolean logic. We investigate whether such a uniform single-operator representation can accelerate either the forward simulation or the parameter identification of a six-branch RC equivalent-circuit model (6rc ECM) of a lithium-ion battery cell. We give the analytical EML rewrite of the discretized state-space recursion, derive an exact operation count, and quantify the depth penalty of the master-formula construction used for gradient-based symbolic regression. Our analysis shows that direct EML simulation is slower than the classical exponential-Euler scheme (a ~ 25x instruction overhead per RC branch), but EML-based parametrization offers a structurally complete, gradient-differentiable basis that competes favourably with non-parametric DRT deconvolution and metaheuristic optimisation when the cardinality of RC branches is unknown a priori. We conclude with a concrete recommendation: use EML only on the parametrization side of the 6rc workflow, keeping the classical recursion at runtime.

Eymen Ipek

Deep neural networks (DNNs) deliver state-of-the-art accuracy on regression and classification tasks, yet two structural deficits persistently obstruct their deployment in safety-critical, resource-constrained settings: (i) opacity of the learned function, which precludes formal verification, and (ii) reliance on heterogeneous, library-bound activation functions that inflate latency and silicon area on edge hardware. The recently introduced Exp-Minus-Log (EML) Sheffer operator, eml(x, y) = exp(x) - ln(y), was shown by Odrzywolek (2026) to be sufficient - together with the constant 1 - to express every standard elementary function as a binary tree of identical nodes. We propose to embed EML primitives inside conventional DNN architectures, yielding a hybrid DNN-EML model in which the trunk learns distributed representations and the head is a depth-bounded, weight-sparse EML tree whose snapped weights collapse to closed-form symbolic sub-expressions. We derive the forward equations, prove computational-cost bounds, analyse inference and training acceleration relative to multilayer perceptrons (MLPs) and physics-informed neural networks (PINNs), and quantify the trade-offs for FPGA/analog deployment. We argue that the DNN-EML pairing closes a literature gap: prior neuro-symbolic and equation-learner approaches (EQL, KAN, AI-Feynman) work with heterogeneous primitive sets and do not exploit a single hardware-realisable Sheffer element. A balanced assessment shows that EML is unlikely to accelerate training, and on commodity CPU/GPU it is also unlikely to accelerate inference; however, on a custom EML cell (FPGA logic block or analog circuit) the asymptotic latency advantage can reach an order of magnitude with simultaneous gain in interpretability and formal-verification tractability.