arXiv subjects

Sauro Succi

Publications and source records attributed to Sauro Succi.

At least 37 records · Page 2Linked to original sources

arXiv2026

Kinetic-based regularization: Learning spatial derivatives and PDE applications

Accurate estimation of spatial derivatives from discrete and noisy data is central to scientific machine learning and numerical solutions of PDEs. We extend kinetic-based regularization (KBR), a localized multidimensional kernel regression method with a single trainable parameter, to learn spatial derivatives with provable second-order accuracy in 1D. Two derivative-learning schemes are proposed: an explicit scheme based on the closed-form prediction expressions, and an implicit scheme that solves a perturbed linear system at the points of interest. The fully localized formulation enables efficient, noise-adaptive derivative estimation without requiring global system solving or heuristic smoothing. Both approaches exhibit quadratic convergence, matching second-order finite difference for clean data, along with a possible high-dimensional formulation. Preliminary results show that coupling KBR with conservative solvers enables stable shock capture in 1D hyperbolic PDEs, acting as a step towards solving PDEs on irregular point clouds in higher dimensions while preserving conservation laws.

math.NA↗

arXiv2026

Variational-Adiabatic Quantum Solver for Systems of Linear Equations with Warm Starts

We propose a revisited variational quantum solver for linear systems, designed to circumvent the barren plateau phenomenon by combining two key techniques: adiabatic evolution and warm starts. To this end, we define an initial Hamiltonian with a known ground state which is easily implemented on the quantum circuit, and then "adiabatically" evolve the Hamiltonian by tuning a control variable in such a way that the final ground state matches the solution to the given linear system. This evolution is carried out in incremental steps, and the ground state at each step is found by minimizing the energy using the parameter values corresponding to the previous minimum as a warm start to guide the search. As a first test case, the method is applied to several linear systems obtained by discretizing a one-dimensional heat flow equation with different physical assumptions and grid choices. Our method successfully and reliably improves upon the solution to the same problem as obtained by a conventional quantum solver, reaching very close to the global minimum also in the case of very shallow circuit implementations.

quant-ph↗

arXiv2026

Generalized Onsager-Regularized Lattice Boltzmann Method for error-free Navier-Stokes models on standard lattices

This work presents a novel strategy to address Navier-Stokes modelling errors arising on first-nearest neighbour lattice Boltzmann (LB) methods and introduces fully local corrections through Onsager-Regularized (OReg) non-equilibrium populations. The proposed mechanism, which admits partially and completely corrected OReg models, is used to develop representative partially and completely corrected models for the six-moment-constrained guided equilibrium (GEq) representation on the D2Q9 lattice. The former realization only addresses compatibility condition violations and improves the accuracy by two/four orders of magnitude at reference/arbitrary lattice temperatures respectively, while the latter additionally corrects stress tensor modelling errors, resulting in a fully corrected exact model. Numerical benchmarks of the corrected schemes demonstrate improved accuracy and stability in comparison to the Lattice-BGK and uncorrected OReg-GEq schemes thus presenting a promising avenue for OReg based thermohydrodynamic extensions.

physics.comp-ph↗

arXiv2026

Block encoding of sparse matrices with a periodic diagonal structure

Block encoding is a successful technique used in several powerful quantum algorithms. In this work we provide an explicit quantum circuit for block encoding a sparse matrix with a periodic diagonal structure. The proposed methodology is based on the linear combination of unitaries (LCU) framework and on an efficient unitary operator used to project the complex exponential at a frequency $ω$ multiplied by the computational basis into its real and imaginary components. We demonstrate a distinct computational advantage with a $\mathcal{O}(\text{poly}(n))$ gate complexity, where $n$ is the number of qubits, in the worst-case scenario used for banded matrices, and $\mathcal{O}(n)$ when dealing with a simple diagonal matrix, compared to the exponential scaling of general-purpose methods for dense matrices. Various applications for the presented methodology are discussed in the context of solving differential problems such as the advection-diffusion-reaction (ADR) dynamics, using quantum algorithms with optimal scaling, e.g., quantum singular value transformation (QSVT). Numerical results are used to validate the analytical formulation.

quant-ph↗

arXiv2026

Spontaneous epicuticular charging affects droplet dynamics on living leaves

How water droplets move and slide on leaves influences plant ecophysiological and abiotic interactions, as well as the design of advanced bio-inspired wetting materials. Despite cross-disciplinary relevance, current descriptions of the in situ dynamics of droplets on living leaves focus almost exclusively on surface structure and chemistry, treating the leaf as a static, electrically neutral substrate. Here, three decades after the mechanistic discovery of the Lotus effect, we show that a yet 'hidden' force due to instantaneous electrical phenomena affect the dynamic droplet motion on living leaves. Using high-speed motion tracking and precision charge measurements, we show that droplets sliding on the pristine epicuticular wax layer on superhydrophobic Colocasia esculenta leaves strongly charge affecting its dynamics, previously observed only on synthetic (highly electronegative fluorinated) surfaces. Droplets accumulate charges of Qp,D1 = -0.02 to -0.15 nC per 30 uL droplet on pristine leaves. However, we specifically demonstrate the crucial role of the epicuticular wax layer plasticity: by a structural modification that decreases its roughness amplitude, the same leaves gain an impressive 30-40 fold enhancement in charge transfer (reaching Qt,D1 = -2.8 to -5.2 nC) slowing the droplet by half due to an estimated electrostatic force of 11 uN dominating the resistive forces. The charge accumulation is surface-history-dependent and charge quantities per droplet are surprisingly similar or even exceeding those recently reported from artificial surfaces. Our findings prove that electrostatic charging is a fundamental component of droplet-leaf interactions, opening new research directions from charge-affected leaf ecology to sustainable materials for droplet-based energy harvesting by tuning surface treatments and, moreover,...

physics.flu-dyn↗

arXiv2026

Ghost-Mode Filtered Fluctuating Lattice Boltzmann Method

Fluctuating lattice Boltzmann solvers are widely employed to model mesoscopic fluid behavior in soft-matter systems, including colloidal suspensions and dilute polymer solutions. Despite their utility, these methods can lose accuracy and stability when non-hydrodynamic modes interfere with the dynamics, especially in single--relaxation-time schemes. Here, we introduce a ghost-mode filtered fluctuating lattice Boltzmann method (GMF-FLBM) for the D3Q27 lattice, obtained by selectively eliminating the propagation of the ghost deterministic content while preserving the necessary stochastic forcing. We show, over a broad range of relaxation times, that GMF-FLBM recovers the amplitudes of equilibrium fluctuations with a comparable accuracy as a fully regularized high-order formulation, while requiring only minor adjustments to the conventional BGK collision framework.

cond-mat.soft↗

arXiv2026

Quantum Algorithm for the Fixed-Radius Neighbor Search

Neighbor search is a computationally demanding problem, usually both time- and memory-consuming. The main problem of this kind of algorithms is the long execution time due to cache misses. In this work, we propose a quantum algorithm for the Fixed RAdius Neighbor Search problem (FRANS) based on the fixed-point version of Grover's algorithm. We propose an efficient circuit for solving the FRANS with linear query complexity with the number of particles $N$. The quantum circuit returns the list of all the neighbors' pairs within the fixed radius, together with their distance, avoiding the slow down given by cache miss. We analyzed the gate and the query complexity of the circuit. Our FRANS algorithm presents a query complexity of $\mathcal{O}(N/\sqrt{M})$, where $M$ is the number of solutions, reaching the optimal lower bound of the Grover's algorithm. We propose different implementations of the oracle, which must be chosen depending on the precise structure of the database. Among these, we present an implementation using the Chebyshev distance with depth $\mathcal{O}(q_1)$, where $2^{q_1}$ is the number of grid points used to discretize a spatial dimension. State-of-the-art algorithms for state preparation allow for a trade-off between depth and width of the circuit, with a volume (depth$\times$ width) of $\mathcal{O}(N\log(N))$. This unfavorable scaling can be brought down to $\mathcal{O}(\text{poly}(\log N))$ in case of structured datasets. We proposed a stopping criterion based on Bayes interference and tested its validity on $1D$ simulations. Finally, we accounted for the readout complexity and assessed the resilience of the model to the readout error, suggesting an error correction-free strategy to check the accuracy of the results.

quant-ph↗

arXiv2025

A new kind of science

We discuss whether science is in the process of being transformed from a quest for causality to a quest for correlation in light of the recent development in artificial intelligence. We observe that while a blind trust in the most seductive promises of AI is surely to be avoided, a judicious combination of computer simulation based on physical insight and the machine learning ability to explore ultra-dimensional spaces, holds potential for transformative progress in the way science is going to be pursued in the years to come.

physics.soc-ph↗

arXiv2025

Regularized Fluctuating Lattice Boltzmann Model

We introduce a regularized fluctuating lattice Boltzmann model (Reg-FLBM) for the D3Q27 lattice, which incorporates thermal fluctuations through Hermite-based projections to ensure compliance with the fluctuation-dissipation theorem. By leveraging the recursive regularization framework, the model achieves thermodynamic consistency for both hydrodynamic and ghost modes. Compared to the conventional single-relaxation-time BGK-FLBM, the Reg-FLBM provides improved stability and a more accurate description of thermal fluctuations. The implementation is optimized for large-scale parallel simulations on GPU-accelerated architectures, enabling systematic investigation of fluctuation-driven phenomena in mesoscale and nanoscale fluid systems.

physics.comp-ph↗

arXiv2025

The foundational value of quantum computing for classical fluids

Quantum algorithms for classical physics problems expose new patterns of quantum information flow as compared to the many-body Schrödinger equation. As a result, besides their potential practical applications, they also offer a valuable theoretical and computational framework to elucidate the foundations of quantum mechanics, particularly the validity of the many-body Schrödinger equation in the limit of large number of particles, on the order of the Avogadro number. This idea is illustrated by means of a concrete example, the Block-Encoded Carleman embedding of the Lattice Boltzmann formulation of fluid dynamics (CLB).

quant-ph↗

arXiv2025

Representation of Typical Droplet Shapes in 2D Latent Space

In this study, we investigate the clustering of 5000 droplets, each originating from one of five distinct droplet classes, each representing a unique geometry. The shape coordinates of the droplets are mapped to a 2D latent space through a two-step, fully reversible process involving Fourier series and autoencoders. Thus, each droplet contour sampled by 400 points is represented by 2 scalar numbers. We present data mapping 5000 droplets in two-dimensional space, which reveals ringlike clusters. Each cluster corresponds to a distinct droplet class, with the classification based on the droplet morphology, which is known a priori. Further, we find that well-known clustering algorithms like DBSCAN show limited success in identifying these ring-shaped clusters underscoring the need to explore alternative clustering approaches, particularly those tailored to handle ring-shaped distributed data.

physics.flu-dyn↗

arXiv2025

A kinetic-based regularization method for data science applications

We propose a physics-based regularization technique for function learning, inspired by statistical mechanics. By drawing an analogy between optimizing the parameters of an interpolator and minimizing the energy of a system, we introduce corrections that impose constraints on the lower-order moments of the data distribution. This minimizes the discrepancy between the discrete and continuum representations of the data, in turn allowing to access more favorable energy landscapes, thus improving the accuracy of the interpolator. Our approach improves performance in both interpolation and regression tasks, even in high-dimensional spaces. Unlike traditional methods, it does not require empirical parameter tuning, making it particularly effective for handling noisy data. We also show that thanks to its local nature, the method offers computational and memory efficiency advantages over Radial Basis Function interpolators, especially for large datasets.

cs.LG↗

arXiv2025

The wall confronting large language models

We show that the scaling laws which determine the performance of large language models (LLMs) severely limit their ability to improve the uncertainty of their predictions. As a result, raising their reliability to meet the standards of scientific inquiry is intractable by any reasonable measure. We argue that the very mechanism which fuels much of the learning power of LLMs, namely the ability to generate non-Gaussian output distributions from Gaussian input ones, might well be at the roots of their propensity to produce error pileup, ensuing information catastrophes and degenerative AI behaviour. This tension between learning and accuracy is a likely candidate mechanism underlying the observed low values of the scaling components. It is substantially compounded by the deluge of spurious correlations pointed out by Calude and Longo which rapidly increase in any data set merely as a function of its size, regardless of its nature. The fact that a degenerative AI pathway is a very probable feature of the LLM landscape does not mean that it must inevitably arise in all future AI research. Its avoidance, which we also discuss in this paper, necessitates putting a much higher premium on insight and understanding of the structural characteristics of the problems being investigated.

cs.AI↗

arXiv2025

Towards nonlinear thermohydrodynamic simulations via the Onsager-Regularized Lattice Boltzmann Method

This work presents a generalized, assumption-free, and stencil-independent theoretical analyses of the recently proposed Onsager-Regularized (OReg) lattice Boltzmann (LB) method [Jonnalagadda et al., Phys. Rev. E 104, 015313 (2021)] and demonstrates its ability to mitigate spurious errors associated with the insufficient isotropy of standard first-neighbor lattices without the inclusion of any external correction terms. The hydrodynamic limit recovered by the OReg scheme is derived for two equilibrium distribution functions, namely the so-called thermal guided equilibrium and the popular second order polynomial equilibrium, to show that the OReg scheme yields macroscopic dynamics that are O(u) times more accurate than that of the bare BGK collision model. Specifically, we show that, with the guided equilibrium on the D2Q9 standard lattice, the OReg scheme inherently compensates for the insufficient lattice isotropy of the standard D2Q9 lattice by automatically adjusting the lattice viscosity, yielding O(u4) and O(u2) accurate kinetic models when operated at the reference and arbitrary temperatures respectively. Further, we also show that the OReg scheme presents an O(u3) accurate kinetic model for the D2Q9 lattice when used with the second order polynomial equilibrium formulation. Thereafter, the accuracy of the OReg-guided-equilibrium kinetic model is numerically demonstrated for quasi-one-dimensional simulations of the rotated decaying shear wave and isothermal shocktube problems. The present work lays the theoretical foundation of a generic framework which can enable fully local, correction-free, nonlinear thermohydrodynamic LB simulations on standard lattices, thereby facilitating scalable simulations of physically challenging fluid flows.

physics.comp-ph↗

arXiv2025

Is the end of Insight in Sight ?

The rise of deep learning challenges the longstanding scientific ideal of insight - the human capacity to understand phenomena by uncovering underlying mechanisms. In many modern applications, accurate predictions no longer require interpretable models, prompting debate about whether explainability is a realistic or even meaningful goal. From our perspective in physics, we examine this tension through a concrete case study: a physics-informed neural network (PINN) trained on a rarefied gas dynamics problem governed by the Boltzmann equation. Despite the system's clear structure and well-understood governing laws, the trained network's weights resemble Gaussian-distributed random matrices, with no evident trace of the physical principles involved. This suggests that deep learning and traditional simulation may follow distinct cognitive paths to the same outcome - one grounded in mechanistic insight, the other in statistical interpolation. Our findings raise critical questions about the limits of explainable AI and whether interpretability can - or should-remain a universal standard in artificial reasoning.

physics.comp-ph↗

arXiv2025

Statistical Characterization of Entanglement Degradation Under Markovian Noise in Composite Quantum Systems

Understanding how noise degrades entanglement is crucial for the development of reliable quantum technologies. While the Markovian approximation simplifies the analysis of noise, it remains computationally demanding, particularly for high-dimensional systems like quantum memories. In this paper, we present a statistical approach to study the impact of different noise models on entanglement in composite quantum systems. By comparing global and local noise scenarios, we quantify entanglement degradation using the Positive Partial Transpose Time (PPTT) metric, which measures how long entanglement persists under noise. When the sampling of different noise scenarios is performed under controlled and homogeneous conditions, our analysis reveals that systems subjected to global noise tend to exhibit longer PPTTs, whereas those influenced by independent local noise models display the shortest entanglement persistence. To carry out this analysis, we employ a computational method proposed by Cao and Lu, which accelerates the simulation of PPTT distributions and enables efficient analysis of systems with dimensions up to $D=8$. Our results demonstrate the effectiveness of this approach for investigating the resilience of quantum systems under Markovian noise.

quant-ph↗

arXiv2025

accLB: A High-Performance Lattice Boltzmann Code for Multiphase Turbulence on Multi-Gpu Architectures

In this work, we present accLB, a high-performance Fortran-based lattice Boltzmann (LB) solver tailored to multiphase turbulent flows on multi-GPU architectures. The code couples a conservative phase-field formulation of the Allen-Cahn equation with a thread-safe, regularized LB method to capture complex interface dynamics. Designed from the ground up for HPC environments, accLB employs MPI for distributed memory parallelism and OpenACC for GPU acceleration, achieving excellent portability and scalability on leading pre-exascale systems such as Leonardo and LUMI. Benchmark tests demonstrate strong and weak scaling efficiencies on multiple GPUs. Physical validation includes direct numerical simulations of homogeneous isotropic turbulence (HIT). Further, we examine bubble-laden HIT and observe a transition to a $-3$ energy scaling, as in experiments and theoretical predictions, due to bubble-induced dissipation, along with enhanced small-scale intermittency. These results highlight accLB as a robust and scalable platform for the simulation of multiphase turbulence in extreme computational regimes.

physics.flu-dyn↗

arXiv2025

Accurately Simulating the Time Evolution of an Ising Model with Echo Verified Clifford Data Regression on a Superconducting Quantum Computer

We present an error mitigation strategy composed of Echo Verification (EV) and Clifford Data Regression (CDR), the combination of which allows one to learn the effect of the quantum noise channel to extract error mitigated estimates for the expectation value of Pauli observables. We analyse the behaviour of the method under the depolarizing channel and derive an estimator for the depolarization rate in terms of the ancilla purity and postselection probability. We also highlight the sensitivity of this probability to noise, a potential bottleneck for the technique. We subsequently consider a more general noise channel consisting of arbitrary Pauli errors, which reveals a linear relationship between the error rates and the estimation of expectation values, suggesting the learnability of noise in EV by regression techniques. Finally, we present a practical demonstration of Echo Verified Clifford Data Regression (EVCDR) on a superconducting quantum computer and observe accurate results for the time evolution of an Ising model over spin-lattices consisting of up to 35 sites and circuit depths up to 173 entangling layers.

quant-ph↗