arXiv subjects

Junyu Zhou

Publications and source records attributed to Junyu Zhou.

At least 19 recordsLinked to original sources

arXiv2026

Reduction Modulo Binary Polynomials with Logarithmic Feedback Depth

Polynomial modular reduction is central to binary finite-field arithmetic and repeated Frobenius powering. Top-down shift/XOR folding can use few operations when the nonleading support is sparse, but a tap near the leading term creates a long feedback chain. We formulate this recurrence as inversion of a nilpotent shift operator and factor its inverse by characteristic-two Frobenius powers. The resulting Frobenius-factorized reduction (FFR) applies to every monic binary modulus without materializing a reciprocal or dense reduction matrix, and its shifts can be generated online without a persistent modulus-specific schedule. For degree $m$ and nonempty nonleading support of size $s$, with nearest-tap distance $\Delta_{\min}$, FFR uses exactly $\lceil\log_2(m/\Delta_{\min})\rceil$ sequential feedback stages and has scheduled work $O(ms(1+\log(m/s)))$. A portable-C evaluation on 1,096 supports through degree $131072$ identifies distinct FFR, L\'opez--Dahab, and gf2x-backed Barrett regions. On four certified irreducible moduli, FFR makes complete Rabin irreducibility testing $1.35$--$8.04$ times faster than NTL and $1.60$--$6.81$ times faster than the matched Barrett implementation.

math.NT↗

arXiv2026

Cyclic Incidence Orderings of Complete Graphs and 3-Uniform Hypergraphs

We study cyclic orderings of all edges of a complete $k$-uniform hypergraph on $n$ vertices in which the binary incidence sequences of the vertices are cyclic shifts of a common word. The shifts are chosen independently, with no prescribed action on the vertices. For $2\leq k<n$, coprimality $\gcd(n,k)=1$ is known to suffice even when consecutive edges must differ by a single vertex exchange. We recall a short orbit construction and prove the converse for the first two nontrivial uniformities without any adjacency requirement. For $k=2$, an ordering exists exactly when $n=2$ or $n$ is odd; for $k=3$, exactly when $n=3$ or $3\nmid n$. The necessity proofs use reflected convolution identities and pair-intersection counts to constrain the vertex shifts to a torsion coset. For triples, multiplicity-preserving dilation and conditional prime-power capacity bounds complete the argument.

math.CO↗

arXiv2026

Optimal Rates for Generalization of Gradient Descent Methods with Deep Neural Networks

Recent progress has been made in understanding the statistical generalization performance of gradient descent methods for overparameterized neural networks within the neural tangent kernel (NTK) regime. However, most of the existing work on regression problems is limited to shallow network architectures, leaving a notable gap in the theory of deep neural networks. This paper addresses this gap by presenting a comprehensive generalization analysis for deep ReLU networks trained using gradient descent (GD) and stochastic gradient descent (SGD). Specifically, we establish the first known minimax-optimal rates of excess population risk for both GD and SGD with deep ReLU networks, under the assumption that the network width scales polynomially with respect to the network depth and training sample size. Our results demonstrate that with sufficient width, gradient descent methods for deep ReLU networks can achieve optimal generalization rates on par with kernel methods.

stat.ML↗

arXiv2026

Minimax-Optimal Generalization Bounds for Smooth Deep Neural Networks Trained by (Stochastic) Gradient Descent

Characterizing the optimization dynamics and statistical performance of over-parameterized deep neural networks (DNNs) remains a central challenge in understanding the remarkable success of deep learning. We establish quantitative bounds showing that kernel gradient descent in the reproducing kernel Hilbert space induced by the deterministic infinite-width neural tangent kernel approximates finite-width deep regression with smooth activations under gradient descent (GD) and stochastic gradient descent (SGD) training. The approximation gap is governed by the network width and training horizon, with an additional stochastic gradient error in the SGD case. This connection provides a general mechanism for transferring learning-theoretic guarantees from kernel methods to deep regression. As an application, under general source and effective dimension conditions, we show that both GD- and SGD-trained DNNs attain the minimax-optimal excess population risk rate, up to logarithmic factors, provided that the network width grows polynomially in the sample size. To the best of our knowledge, these are the first such guarantees for standard fully connected deep neural networks with smooth activations trained by GD and SGD.

stat.ML↗

arXiv2026

GeneCS: Synthesizing Resource-Efficient Code Surgery for Arbitrary Quantum Stabilizer Codes

Efficiently realizing logical operations on general stabilizer codes remains a long-standing challenge in fault tolerant quantum computing. While code surgery provides a general framework with provable guarantees by joint logical measurements, existing constructions are largely theoretical and incur substantial ancilla overhead in practice. In this work, we propose GeneCS, a resource-efficient compiler for synthesizing code surgery protocols for arbitrary stabilizer codes. Our approach leverages structure-aware optimizations to eliminate redundancy in graph construction, dynamically balance expansion and congestion, and incorporate code degree constraints. Experimental results show that GeneCS achieves an average reduction of over $85\%$ in ancillary qubits and checks for both single-code and cross-code logical operations, while preserving logical error rates. Moreover, our compiler scales to codes with more than $10^4$ qubits with an amortized compilation time of about one second per instance. These results enable practical logical operations and efficient cross-code communication, thereby supporting the deployment of modern QLDPC codes and heterogeneous quantum architectures.

quant-ph↗

arXiv2026

Population Risk Bounds for Kolmogorov-Arnold Networks Trained by DP-SGD with Correlated Noise

We establish the first population risk bounds for Kolmogorov-Arnold Networks (KANs) trained by mini-batch SGD with gradient clipping, covering non-private SGD as well as differentially private SGD (DP-SGD) with Gaussian perturbations that interpolate between independent and temporally correlated noise. This setting is substantially closer to practice than prior KAN theory along two axes: training is by mini-batch SGD, the standard recipe for modern networks, rather than full-batch gradient descent (GD); and correlated-noise mechanisms have empirically shown a more favorable privacy-utility tradeoff than independent-noise mechanisms. Our results cover the corresponding full-batch GD and independent-noise DP-GD results for KANs by Wang et al. (2026), while yielding sharper fixed-second-layer specializations. The technical core is a new analysis route for correlated-noise DP training in the non-convex regime. Temporal dependence breaks the conditional-centering structure underlying standard one-step SGD arguments, and the projection step obstructs the exact cancellation structure of correlated perturbations. We address these difficulties through an auxiliary unprojected dynamics, a shifted iterate that absorbs the current noise perturbation, and a high-probability bootstrap certifying projection inactivity. Combining this optimization analysis with a stability-based generalization argument yields the stated population risk bounds. To the best of our knowledge, this is the first optimization and population risk analysis of a correlated-noise mechanism for DP training beyond convex learning, in particular for neural networks.

cs.LG↗

arXiv2026

TopoLS: Lattice Surgery Compilation via Topological Program Transformations

Lattice surgery is a leading approach for implementing fault-tolerant logical operations in surface code quantum computing, but compiling efficient lattice surgery layouts remains challenging. Existing compilers are largely circuit-centric and operate directly on gate sequences, limiting their ability to exploit the topological flexibility of merge-split operations and minimize space--time volume. We present TopoLS, a topology-centric compiler that uses ZX diagrams as an intermediate representation for lattice surgery compilation. TopoLS combines semantic-preserving ZX-level program transformations, including spider fusion and topology-aware slicing, with a Monte Carlo Tree Search (MCTS)-based synthesis procedure that constructs pipe-diagram embeddings by jointly optimizing placement and routing in 3D space--time. To scale to large circuits, TopoLS further introduces topology-aware partitioning that decomposes the compilation task into bounded subproblems and limits the routing frontier during embedding. Across evaluated benchmarks, TopoLS achieves an average $46\%$ reduction in space--time volume over prior circuit-centric compilers, with improvements ranging from $25\%$ to $90\%$, and exhibits strong empirical scalability on large benchmark families. Compared with SAT-based formulations that become intractable on larger instances, TopoLS offers a practical end-to-end solution for optimized lattice surgery compilation. TopoLS has been integrated into the TQEC ecosystem, enabling downstream circuit-level simulation and resource estimation workflows.

quant-ph↗

arXiv2026

Optimization, Generalization and Differential Privacy Bounds for Gradient Descent on Kolmogorov-Arnold Networks

Kolmogorov--Arnold Networks (KANs) have recently emerged as a structured alternative to standard MLPs, yet a principled theory for their training dynamics, generalization, and privacy properties remains limited. In this paper, we analyze gradient descent (GD) for training two-layer KANs and derive general bounds that characterize their training dynamics, generalization, and utility under differential privacy (DP). As a concrete instantiation, we specialize our analysis to logistic loss under an NTK-separable assumption, where we show that polylogarithmic network width suffices for GD to achieve an optimization rate of order $1/T$ and a generalization rate of order $1/n$, with $T$ denoting the number of GD iterations and $n$ the sample size. In the private setting, we characterize the noise required for $(\epsilon,\delta)$-DP and obtain a utility bound of order $\sqrt{d}/(n\epsilon)$ (with $d$ the input dimension), matching the classical lower bound for general convex Lipschitz problems. Our results imply that polylogarithmic width is not only sufficient but also necessary under differential privacy, revealing a qualitative gap between non-private (sufficiency only) and private (necessity also emerges) training regimes. Experiments further illustrate how these theoretical insights can guide practical choices, including network width selection and early stopping.

cs.LG↗

arXiv2026

AlphaSyndrome: Tackling the Syndrome Measurement Circuit Scheduling Problem for QEC Codes

Quantum error correction (QEC) is essential for scalable quantum computing, yet repeated syndrome-measurement cycles dominate its spacetime and hardware cost. Although stabilizers commute and admit many valid execution orders, different schedules induce distinct error-propagation paths under realistic noise, leading to large variations in logical error rate. Outside of surface codes, effective syndrome-measurement scheduling remains largely unexplored. We present AlphaSyndrome, an automated synthesis framework for scheduling syndrome-measurement circuits in general commuting-stabilizer codes under minimal assumptions: mutually commuting stabilizers and a heuristic decoder. AlphaSyndrome formulates scheduling as an optimization problem that shapes error propagation to (i) avoid patterns close to logical operators and (ii) remain within the decoder's correctable region. The framework uses Monte Carlo Tree Search (MCTS) to explore ordering and parallelism, guided by code structure and decoder feedback. Across diverse code families, sizes, and decoders, AlphaSyndrome reduces logical error rates by 80.6% on average (up to 96.2%) relative to depth-optimal baselines, matches Google's hand-crafted surface-code schedules, and outperforms IBM's schedule for the Bivariate Bicycle code.

cs.ET↗

arXiv2025

TeleWorld: Towards Dynamic Multimodal Synthesis with a 4D World Model

World models aim to endow AI systems with the ability to represent, generate, and interact with dynamic environments in a coherent and temporally consistent manner. While recent video generation models have demonstrated impressive visual quality, they remain limited in real-time interaction, long-horizon consistency, and persistent memory of dynamic scenes, hindering their evolution into practical world models. In this report, we present TeleWorld, a real-time multimodal 4D world modeling framework that unifies video generation, dynamic scene reconstruction, and long-term world memory within a closed-loop system. TeleWorld introduces a novel generation-reconstruction-guidance paradigm, where generated video streams are continuously reconstructed into a dynamic 4D spatio-temporal representation, which in turn guides subsequent generation to maintain spatial, temporal, and physical consistency. To support long-horizon generation with low latency, we employ an autoregressive diffusion-based video model enhanced with Macro-from-Micro Planning (MMPL)--a hierarchical planning method that reduces error accumulation from frame-level to segment-level-alongside efficient Distribution Matching Distillation (DMD), enabling real-time synthesis under practical computational budgets. Our approach achieves seamless integration of dynamic object modeling and static scene representation within a unified 4D framework, advancing world models toward practical, interactive, and computationally accessible systems. Extensive experiments demonstrate that TeleWorld achieves strong performance in both static and dynamic world understanding, long-term consistency, and real-time generation efficiency, positioning it as a practical step toward interactive, memory-enabled world models for multimodal generation and embodied intelligence.

cs.CV↗

arXiv2025

F2: Offline Reinforcement Learning for Hamiltonian Simulation via Free-Fermionic Subroutine Compilation

Compiling shallow and accurate quantum circuits for Hamiltonian simulation remains challenging due to hardware constraints and the combinatorial complexity of minimizing gate count and circuit depth. Existing optimization method pipelines rely on hand-engineered classical heuristics, which cannot learn input-dependent structure and therefore miss substantial opportunities for circuit reduction. We introduce F2, an offline reinforcement learning framework that exploits free-fermionic structure to efficiently compile Trotter-based Hamiltonian simulation circuits. F2 provides (i) a reinforcement-learning environment over classically simulatable free-fermionic subroutines, (ii) architectural and objective-level inductive biases that stabilize long-horizon value learning, and (iii) a reversible synthetic-trajectory generation mechanism that consistently yields abundant, guaranteed-successful offline data. Across benchmarks spanning lattice models, protein fragments, and crystalline materials (12-222 qubits), F2 reduces gate count by 47% and depth by 38% on average relative to strong baselines (Qiskit, Cirq/OpenFermion) while maintaining average errors of 10^(-7). These results show that aligning deep reinforcement learning with the algebraic structure of quantum dynamics enables substantial improvements in circuit synthesis, suggesting a promising direction for scalable, learning-based quantum compilation

quant-ph↗

arXiv2025

Diffusion-Driven Progressive Target Manipulation for Source-Free Domain Adaptation

Source-free domain adaptation (SFDA) is a challenging task that tackles domain shifts using only a pre-trained source model and unlabeled target data. Existing SFDA methods are restricted by the fundamental limitation of source-target domain discrepancy. Non-generation SFDA methods suffer from unreliable pseudo-labels in challenging scenarios with large domain discrepancies, while generation-based SFDA methods are evidently degraded due to enlarged domain discrepancies in creating pseudo-source data. To address this limitation, we propose a novel generation-based framework named Diffusion-Driven Progressive Target Manipulation (DPTM) that leverages unlabeled target data as references to reliably generate and progressively refine a pseudo-target domain for SFDA. Specifically, we divide the target samples into a trust set and a non-trust set based on the reliability of pseudo-labels to sufficiently and reliably exploit their information. For samples from the non-trust set, we develop a manipulation strategy to semantically transform them into the newly assigned categories, while simultaneously maintaining them in the target distribution via a latent diffusion model. Furthermore, we design a progressive refinement mechanism that progressively reduces the domain discrepancy between the pseudo-target domain and the real target domain via iterative refinement. Experimental results demonstrate that DPTM outperforms existing methods by a large margin and achieves state-of-the-art performance on four prevailing SFDA benchmark datasets with different scales. Remarkably, DPTM can significantly enhance the performance by up to 18.6% in scenarios with large source-target gaps.

cs.CV↗

arXiv2025

NeuroQD: A Learning-Based Simulation Framework For Quantum Dot Devices

Electron spin qubits in quantum dot devices are promising for scalable quantum computing. However, architectural support is currently hindered by the lack of realistic and performant simulation methods for real devices. Physics-based tools are accurate yet too slow for simulating device behavior in real-time, while qualitative models miss layout and wafer heterostructure. We propose a new simulation approach capable of simulating real devices from the cold-start with real-time performance. Leveraging a key phenomenon observed in physics-based simulation, we train a compact convolutional neural network (CNN) to infer the qubit-layer electrostatic potential from gate voltages. Our GPU-accelerated inference delivers >1000x speedup with >96% agreement to the physics-based simulation. Integrated into the experiment control stack, the simulator returns results with millisecond scale latency, reproduces key tuning features, and yields device behaviors and metrics consistent with measurements on devices operated at 9 mK.

cond-mat.mes-hall↗

arXiv2025

3DGabSplat: 3D Gabor Splatting for Frequency-adaptive Radiance Field Rendering

Recent prominence in 3D Gaussian Splatting (3DGS) has enabled real-time rendering while maintaining high-fidelity novel view synthesis. However, 3DGS resorts to the Gaussian function that is low-pass by nature and is restricted in representing high-frequency details in 3D scenes. Moreover, it causes redundant primitives with degraded training and rendering efficiency and excessive memory overhead. To overcome these limitations, we propose 3D Gabor Splatting (3DGabSplat) that leverages a novel 3D Gabor-based primitive with multiple directional 3D frequency responses for radiance field representation supervised by multi-view images. The proposed 3D Gabor-based primitive forms a filter bank incorporating multiple 3D Gabor kernels at different frequencies to enhance flexibility and efficiency in capturing fine 3D details. Furthermore, to achieve novel view rendering, an efficient CUDA-based rasterizer is developed to project the multiple directional 3D frequency components characterized by 3D Gabor-based primitives onto the 2D image plane, and a frequency-adaptive mechanism is presented for adaptive joint optimization of primitives. 3DGabSplat is scalable to be a plug-and-play kernel for seamless integration into existing 3DGS paradigms to enhance both efficiency and quality of novel view synthesis. Extensive experiments demonstrate that 3DGabSplat outperforms 3DGS and its variants using alternative primitives, and achieves state-of-the-art rendering quality across both real-world and synthetic scenes. Remarkably, we achieve up to 1.35 dB PSNR gain over 3DGS with simultaneously reduced number of primitives and memory consumption.

cs.CV↗

arXiv2025

QTurbo: A Robust and Efficient Compiler for Analog Quantum Simulation

Analog quantum simulation leverages native hardware dynamics to emulate complex quantum systems with great efficiency by bypassing the quantum circuit abstraction. However, conventional compilation methods for analog simulators are typically labor-intensive, prone to errors, and computationally demanding. This paper introduces QTurbo, a powerful analog quantum simulation compiler designed to significantly enhance compilation efficiency and optimize hardware execution time. By generating precise and noiseresilient pulse schedules, our approach ensures greater accuracy and reliability, outperforming the existing state-of-theart approach.

quant-ph↗

arXiv2025

Infant Cry Emotion Recognition Using Improved ECAPA-TDNN with Multiscale Feature Fusion and Attention Enhancement

Infant cry emotion recognition is crucial for parenting and medical applications. It faces many challenges, such as subtle emotional variations, noise interference, and limited data. The existing methods lack the ability to effectively integrate multi-scale features and temporal-frequency relationships. In this study, we propose a method for infant cry emotion recognition using an improved Emphasized Channel Attention, Propagation and Aggregation in Time Delay Neural Network (ECAPA-TDNN) with both multi-scale feature fusion and attention enhancement. Experiments on a public dataset show that the proposed method achieves accuracy of 82.20%, number of parameters of 1.43 MB and FLOPs of 0.32 Giga. Moreover, our method has advantage over the baseline methods in terms of accuracy. The code is at https://github.com/kkpretend/IETMA.

eess.AS↗

arXiv2025

Kernpiler: Compiler Optimization for Quantum Hamiltonian Simulation with Partial Trotterization

Quantum computing promises transformative impacts in simulating Hamiltonian dynamics, essential for studying physical systems inaccessible by classical computing. However, existing compilation techniques for Hamiltonian simulation, in particular the commonly used Trotter formulas struggle to provide gate counts feasible on current quantum computers for beyond-classical simulations. We propose partial Trotterization, where sets of non-commuting Hamiltonian terms are directly compiled allowing for less error per Trotter step and therefore a reduction of Trotter steps overall. Furthermore, a suite of novel optimizations are introduced which complement the new partial Trotterization technique, including reinforcement learning for complex unitary decompositions and high level Hamiltonian analysis for unitary reduction. We demonstrate with numerical simulations across spin and fermionic Hamiltonians that compared to state of the art methods such as Qiskit's Rustiq and Qiskit's Paulievolutiongate, our novel compiler presents up to 10x gate and depth count reductions.

quant-ph↗

arXiv2024

MarQSim: Reconciling Determinism and Randomness in Compiler Optimization for Quantum Simulation

Quantum simulation, fundamental in quantum algorithm design, extends far beyond its foundational roots, powering diverse quantum computing applications. However, optimizing the compilation of quantum Hamiltonian simulation poses significant challenges. Existing approaches fall short in reconciling deterministic and randomized compilation, lack appropriate intermediate representations, and struggle to guarantee correctness. Addressing these challenges, we present MarQSim, a novel compilation framework. MarQSim leverages a Markov chain-based approach, encapsulated in the Hamiltonian Term Transition Graph, adeptly reconciling deterministic and randomized compilation benefits. We rigorously prove its algorithmic efficiency and correctness criteria. Furthermore, we formulate a Min-Cost Flow model that can tune transition matrices to enforce correctness while accommodating various optimization objectives. Experimental results demonstrate MarQSim's superiority in generating more efficient quantum circuits for simulating various quantum Hamiltonians while maintaining precision.

quant-ph↗