Search arXiv⌕ Search

arXiv subjects

Peng Yang

Publications and source records attributed to Peng Yang.

At least 37 records · Page 2Linked to original sources
arXiv2026

StepAudio 3 Gen Technical Report

We introduce StepAudio 3 Gen, a general-purpose audio generation model that supports zero-shot text-to-speech (TTS), voice design, vocal generation, sound effects, music, vibe speech, and mixtures of multiple audio types within a unified framework. At its core, StepAudio 3 Gen is a discrete autoregressive generator that models audio directly over residual vector quantization (RVQ) tokens, departing from the diffusion Transformer-based continuous generation paradigm prevalent in recent general audio models. Its StepAudio Tokenizer represents general audio at 12.5 Hz in a shared $16 \times 2048$ residual code space, jointly quantizing semantic and waveform-level acoustic features so that each code layer preserves both types of information. For generation, the backbone predicts the first codebook along the time axis using autoregressive modeling, while a lightweight causal Transformer completes the remaining fifteen codebooks along the codebook axis. Our study further identifies three key design principles: (1) interference-aware progressive pretraining for acquiring audio capabilities while preserving the textual abilities of the large language model, (2) RVQ Adaptor for effectively incorporating multi-codebook acoustic representations, and (3) discrete autoregressive modeling over a shared representation across general audio domains. With progressive pretraining, multi-task instruction training, and supervised fine-tuning, StepAudio 3 Gen achieves state-of-the-art performance on both TTS and voice design, while retaining strong generation capabilities across speech, vocals, sound effects, and music. Audio samples are available at https://stepaudiollm.github.io/step-audio-3-gen/.

cs.SD↗
arXiv2026

Online Regime-aware Calibration for Black-box Social Simulators via Posterior-assisted Evolutionary Dynamic Optimization

Evolutionary dynamic optimization (EDO) commonly assumes that environmental changes can be detected from fitness variations and handled through random re-initialization, historical solutions, or learned transition patterns. Online calibration of black-box simulators introduces a different setting, where the dynamic objective is induced by sequential observations and a changing calibration window, rather than being controlled by explicit variables. Fitness variations therefore cannot be directly attributed to regime changes, while the unknown relationship between successive regimes limits conventional adaptation. We formulate this setting as an observation-driven dynamic optimization problem and propose PosEDO, which augments fitness-based EDO with an observation-conditioned parameter-space signal. PosEDO learns this signal online as a posterior distribution over simulator parameters from parameter-trajectory pairs generated during evolutionary evaluation, using posterior shifts for change detection and posterior samples for population adaptation. The new evaluation records are further utilized for online posterior updating without additional simulator calls. Experiments on nonstationary economic and financial simulators show that PosEDO improves calibration accuracy, optimization performance, and change-detection quality over representative EDO baselines.

cs.NE↗
arXiv2026

Decentralized and Equilibrium-Set-Oriented Stability Analysis and Control for Power Systems

Conventional power-system stability analysis is largely centralized and centered on a single equilibrium point, which becomes increasingly restrictive in the presence of large-scale fluctuating renewable generation. This paper develops a decentralized framework for stability analysis and control that certifies the asymptotic stability of an equilibrium set rather than that of a given single operating point. To this end, we introduce a new notion termed input--output differential passivity (IODP), which decomposes equilibrium-set stability of the interconnected system into local requirements imposed on individual devices. These requirements are formulated without embedding a particular operating equilibrium into the local conditions; once the certified regions are constructed, stability verification for a given operating scenario reduces to checking whether its equilibrium lies in the certified set. The proposed conditions require each bus to possess a sufficient level of IODP, quantified by an IODP index. To compensate for an IODP shortage, we further develop an I/O-transformation-based passivation controller that reshapes the local input--output behavior of the corresponding device. In this way, all grid-connected components can be made to satisfy the decentralized conditions for system-wide stability. The proposed framework is validated on a modified IEEE 39-bus system. Simulation results demonstrate that it provides a scalable and equilibrium-set-oriented solution for stability certification and control under highly variable operating conditions.

eess.SY↗
arXiv2026

Efficient Parameter Calibration of Numerical Weather Prediction Models via Evolutionary Sequential Transfer Optimization

The configuration of physical parameterization schemes in Numerical Weather Prediction (NWP) models plays a critical role in determining the accuracy of the forecast. However, existing parameter calibration methods typically treat each calibration task as an isolated optimization problem. This approach suffers from prohibitive computational costs and necessitates performing iterative searches from scratch for each task, leading to low efficiency in sequential calibration scenarios. To address this issue, we propose the SEquential Evolutionary Transfer Optimization (SEETO) algorithm driven by the representations of the meteorological state. First, to accurately measure the physical similarity between calibration tasks, a meteorological state representation extractor is introduced to map high-dimensional meteorological fields into latent representations. Second, given the similarity in the latent space, a bi-level adaptive knowledge transfer mechanism is designed. At the solution level, superior populations from similar historical tasks are reused to achieve a "warm start" for optimization. At the model level, an ensemble surrogate model based on source task data is constructed to assist the search, employing an adaptive weighting mechanism to dynamically balance the contributions of source domain knowledge and target domain data. Experiments on multiple calibration tasks with varying source--target similarities demonstrate that SEETO consistently improves early-stage calibration efficiency under limited expensive evaluation budgets, while maintaining competitive overall optimization performance. This provides a practical approach for efficient automated calibration of NWP model parameters.

cs.CE↗
arXiv2026

On the Formality of Configuration Spaces of $\mathbb{R}^{n'} \times \mathbb{C}^{n}$

This paper presents a complete classification of the formality of configuration spaces of $\mathbb{R}^{n'} \times \mathbb{C}^{n}$. We define a constructible de Rham-Dolbeault cohomology theory which provides a constructible CDGA (commutative differential graded algebra) model of $\Conf_m(\mathbb{R}^{n'} \times \mathbb{C}^{n})$. For $(n'=0,n\ge2)$ or $(n'=1,n\ge1)$, the CDGAs are non-formal. For $n'\ge2,n\ge1$, we establish an explicit quasi-isomorphism between the constructible CDGA and its cohomology by using a diagrammatic CDGA of admissible diagrams and a regularized configuration space integral, which leads to the formality. As an application, we show that the local operator algebra of a topological-holomorphic field theory on $\mathbb{R}^{n'} \times \mathbb{C}^{n}$ ($n'\ge2,n\ge1$) is homotopically equivalent to a higher dimensional analog of vertex algebras.

math.AT↗
arXiv2026

Natural superconvergence points for splines

This paper develops a unified theory of natural superconvergence points for polynomial spline approximations to second-order elliptic problems. Beginning with the one-dimensional case, we establish that when a point $x_0$ is a local symmetric center of the partition, the numerical error $(u-u_h)^{(s)}(x_0)$ exhibits superconvergence whenever the polynomial degree $k$ and the derivative order $s$ share the same parity. In particular, for the smoothest spline (B-spline) solution, the abundance of superconvergence points allows us to construct asymptotic expansion of the error within the element that fully characterize all superconvergence points, for both function values and derivatives. The theoretical framework is then extended to higher-dimensional settings on simplicial and tensor-product meshes, and the essential conclusions are preserved, with one-dimensional derivatives generalized to mixed derivatives. Numerical experiments demonstrate that superconvergence persists even in extremely localized symmetric regions, revealing that superconvergence points are both readily attainable and follow systematic distribution patterns.

math.NA↗
arXiv2026

Natural superconvergence points and asymptotic expansions for spline finite elements in one dimension

We study the natural superconvergence points and asymptotic expansions of one-dimensional spline finite element approximations. For a spline space of degree $k$ and any smoothness $0\leμ\le k-1$, we prove that the $s$-th derivative of the error exhibits enhanced convergence of order $O(h^{k+2-s})$ at points where $k-s$ is even, provided the mesh is symmetric within a region of size $Ch|\ln h|$ around the point. This condition is known to be optimal for the cases of low derivative order $s=0,1$; the present analysis shows that the same local condition is sufficient for all admissible $s$. Moreover, by expanding the error in Legendre polynomials, a closure theorem determines the leading-order Legendre coefficients (the asymptotic expansion of the error) by combining the Galerkin orthogonality with the superconvergence conditions. For $μ=k-1$ (B-splines) and $μ=k-2$, the Galerkin orthogonality conditions vanish and the coefficients are determined solely by the superconvergence conditions. The asymptotic expansion can be expressed through a simple antiderivative recurrence on Legendre polynomials. The resulting polynomial's zeros encode the complete set of superconvergence points for all derivative orders. Numerical experiments for selected $(k,μ)$ pairs confirm the theoretical predictions.

math.NA↗
arXiv2026

KC-BFPRL: Knowledge-Guided Multi-UAV Collaboration for Grassland Restoration via Bilevel Formerpointer-Based Reinforcement Learning

Multi-unmanned aerial vehicle (UAV) systems provide scalable service platforms for large-scale environmental tasks, such as grassland ecosystem restoration. However, coordinating fleet operations requires solving the restoration area maximization problem (RAMP). This non-linear combinatorial optimization challenge is complicated by payload-dependent energy dynamics and heterogeneous ecological degradation. We propose a novel knowledge-guided collaborative bilevel formerpointer reinforcement learning framework (KC-BFPRL) to address this complexity. Using a hierarchical paradigm, KC-BFPRL decomposes RAMP into global task allocation and local restoration planning, with the latter further divided into upper-level trajectory planning and lower-level restoration area allocation. Our specialized architecture pairs featuring a Transformer-based encoder that fuses static environmental features with dynamic UAV states, and a Pointer Network decoder trained via a robust actor-critic framework. By embedding ecological priority rules and heuristic logic, KC-BFPRL achieves a structured warm-start, solving the RL cold-start problem while ensuring strict constraint satisfaction. Extensive experiments demonstrate that KC-BFPRL consistently outperforms state-of-the-art baselines, achieving superior objective values and efficiency. It maintains a $0.00\%$ optimality gap in the most complex scenarios U8-R160 and operates nearly three times faster than MAPDP, validating its robustness, scalability, and real-time applicability for large-scale automated ecological restoration.

cs.MA↗
arXiv2026

High-order complete flux schemes for convection-diffusion equations on arbitrary subdivisions

We develop a novel complete flux finite volume method for convection-diffusion equations on arbitrary subdivisions in two and three dimensions. Unlike standard finite volume discretizations, where the numerical flux is directly approximated from the flux definition, we derive the exact normal flux across each control volume edge/face from the underlying PDE. This exact flux splits naturally into a homogeneous part (the classical Scharfetter--Gummel flux) and an inhomogeneous part based on a Green's function that incorporates the tangential flux and the source term. The resulting formulation is exactly equivalent to the continuous equation and, once the discrete space is chosen, yields high-order schemes without using correction or stabilization strategies. From this framework, we develop concrete numerical schemes on arbitrary grids using Lagrange finite element spaces and B-spline spaces, together with their companion dual meshes (control volume partitions). Numerical experiments in two and three dimensions confirm the optimal convergence and positivity preservation of the proposed schemes.

math.NA↗
arXiv2026

FreCast: Refining Radar Echo Intensity via Phase-Preserving Amplitude Residual Diffusion for Precipitation Nowcasting

Precipitation nowcasting predicts the spatiotemporal evolution of future radar echoes from historical radar echo sequences, thereby estimating the occurrence, development, and movement of precipitation over the near term. In recent years, deep learning has become an important approach to precipitation nowcasting. Although state-of-the-art models can generally capture the overall spatial distribution of future precipitation, their predictions still exhibit substantial biases in radar echo intensity at individual locations. This observation motivates a more targeted strategy for reducing forecast errors. Instead of regenerating an entire radar echo sequence without spatial constraints, the predicted precipitation structure can be used to guide the refinement of echo intensities at individual locations. This structure-guided refinement directly targets echo intensity biases. Accordingly, we propose FreCast, a two-stage framework for radar echo prediction. The first stage generates an initial forecast of future radar echoes. The second stage uses the spatial structure of the initial forecast as a constraint to further correct intensity biases at individual locations in the first-stage prediction. Experiments on three datasets demonstrate that FreCast achieves consistent improvements across forecast skill metrics. Qualitative results further show that FreCast better preserves rainband continuity and intense precipitation structures at longer lead times.

cs.CV↗
arXiv2026

Quasinormal Modes of Extremal Reissner-Nordstrom Black Holes via Seiberg-Witten Quantization

We study the scalar perturbations of asymptotically flat extremal Reissner-Nordström black holes via the quantum Seiberg-Witten geometry of $\mathcal{N}=2$ SU(2) gauge theory with $N_f=2$ flavors. The radial master equation, governed by a double confluent Heun equation, is exactly mapped to the quantum Seiberg-Witten curve, providing an exact quantization condition derived from the non-perturbative Nekrasov-Shatashvili free energy. Analytically, this exact dictionary unveils precise gauge-theoretic interpretations for critical physical thresholds, demonstrating that the superradiance and mass decoupling limits naturally reduce the master equation to the Whittaker equation and the reduced doubly confluent Heun equation (the latter corresponds to the SW geometry of the $\mathcal{N}=2$ SU(2) gauge theory with $N_f=1$), respectively. At the strict extremal limit, the coalescence of horizons induces a topological singularity that complicates the spectral analysis. By accommodating this irregular singularity, our geometric framework resolves the singularity coalescence and enables the extraction of the discrete global quasinormal mode. As our main contribution, we provide the first non-perturbative evaluation of the quasinormal modes spectrum for simultaneously charged and massive scalar fields directly at strict extremity. Furthermore, our analytical results reproduce numerical benchmarks for both neutral and charged massless probes, and naturally capture quasi-resonance behaviors.

hep-th↗
arXiv2026

Rare Events Govern Defect Formation under Weak Symmetry Breaking

Crossing a continuous phase transition out of equilibrium typically generates topological defects whose density obeys a universal power-law scaling predicted by the Kibble-Zurek mechanism. Recent numerical studies have revealed systematic deviations from this scaling in the presence of weak explicit symmetry breaking, manifested as an additional exponential suppression of defect formation. However, the origin of this correction and a general theoretical framework to describe it have remained elusive. Here, using large-deviation theory, we show that defect formation under weak symmetry breaking is controlled by rare fluctuations that drive local regions into the disfavored symmetry-broken state. This mechanism yields a closed-form expression for the defect density in arbitrary dimensions, valid in the weak-field and weak-noise limits. These theoretical predictions are verified through direct simulations of stochastic Ginzburg-Landau models in one and two spatial dimensions.

cond-mat.stat-mech↗
arXiv2026

RiverONE: Generating Knowledge-Intensive VLM by Simulated Quantum Machines

Quantum computing provides a powerful paradigm for representing and transforming high-dimensional information through superposition, entanglement, and measurement-induced nonlinear features. While current quantum hardware is not yet practical for direct large-scale vision-language model (VLM) inference, simulated quantum computation can be used during model construction to generate structured parameters for compact classical AI systems. We build RiverONE, a lightweight vision-language model for quantum calibration plot understanding, using simulated quantum computation. It employs a specialized visual encoder and an InternVL-based language backbone. To compensate for compression-induced information loss, we introduce quantum-generated parameters, which are materialized as classical tensors after training. This allows RiverONE to run entirely on classical GPUs at inference time, with no quantum hardware or runtime quantum simulation. With approximately 1.9 billion parameters, RiverONE achieves at least 95\% of the performance of NVIDIA Ising Calibration 1 on quantum calibration plot understanding tasks while using less than 10\% of its parameter count. These results suggest that simulated quantum computation can serve as a practical construction-stage mechanism for building lightweight, knowledge-intensive scientific VLMs. Our code is available at https://github.com/THeWakeSystems/RiverOne.

quant-ph↗
arXiv2026

Adaptive Plug-and-Play Channel Estimation with Consistency Models for MIMO Systems

This paper proposes a consistency-model-based channel estimation algorithm for multiple-input multiple-output (MIMO) systems. The proposed algorithm employs a consistency model (CM) to learn the angle-domain channel distribution and uses the trained CM as a plug-and-play (PnP) generative prior for MIMO channel estimation. The proposed algorithm alternates between a pilot-observation-based data-consistency update and a CM-prior-based denoising update. In addition, the proposed algorithm adaptively selects the penalty parameter according to residual energy and residual whiteness, and adjusts the CM denoising level according to the observed signal-to-noise ratio (SNR), thereby avoiding the performance degradation caused by fixed inference schedules under varying observation conditions. Simulation results show that the proposed algorithm not only reduces the number of inference steps by 50%--90, but also achieves high estimation accuracy and favorable cross-dataset performance.

eess.SP↗
arXiv2026

From Flat to Hierarchical: Evolving Tree-structured Thoughts for Fine-grained Alpha Mining

Alpha mining, aimed at discovering predictive return signals, is typically formulated as symbolic regression. Traditional symbolic methods suffer from search inefficiency and biased prior knowledge. Recently, Large Language Models (LLMs) have emerged as a promising alternative, automatically generating textual thoughts and executable codes to achieve both efficient and interpretable alpha mining. However, existing approaches mostly focus on leveraging LLM's reasoning and reflection capabilities, yet largely neglect the positional bias due to the flat thought representation which restricts efficiency and diversity of the search process. This paper introduces Tree-structured thought Evolution (TreEvo), which evolves hierarchically decomposed thoughts to expand the effective search space. In addition, we propose a set of evolutionary operators tailored to structured thoughts. Experiments on four real-market datasets demonstrate that TreEvo not only obtains competitive alphas with traditional methods in up to 200 times fewer evaluations, but also consistently outperforms LLM-driven EAs across all datasets by $14.31\%$ on average.

cs.CE↗
arXiv2026

StepAudio 2.5 Technical Report

Unified audio-language modeling has emerged as a prominent trend in modern speech systems, promising to bring the reasoning capabilities of large language models to auditory tasks. However, existing unified foundations often struggle to match the depth of specialized systems across automatic speech recognition (ASR), text-to-speech synthesis (TTS), and realtime spoken interaction. Bridging this gap remains an open challenge. This report presents StepAudio 2.5, a unified audio-language foundation model that matches or exceeds specialized systems across all three capabilities. Rather than treating these tasks as architecturally distinct, we operate on the premise that once text and audio share a multimodal representational space, task specialization becomes a matter of operational regimes: data construction, optimization targets, and decoding constraints. Guided by this insight, we advance the post-training paradigm from standard supervised learning to task-tailored Reinforcement Learning from Human Feedback (RLHF), using it as the primary mechanism to define complex optimization targets. We leverage this RLHF-centric alignment, alongside specialized decoding, to shape a shared backbone into three distinct operational modes. Concretely, the ASR branch advances transcription efficiency via verifiable multi-token decoding; the TTS branch achieves controllable, expressive synthesis through preference-based RLHF and context-rich supervision; and the Realtime branch realizes low-latency, persona-consistent dialogue via generative reward modeling within an RLHF framework. On standard benchmarks, StepAudio 2.5 achieves state-of-the-art results across ASR, TTS, and Realtime, demonstrating that a singular audio-language foundation can successfully internalize the distinct deployment objectives of speech understanding, generation, and live interaction.

eess.AS↗
arXiv2026

Omni-scale Learning-based Sequential Decision Framework for Order Fulfillment of Tote-handling Robotic Systems

Driven by the rapid expansion of e-commerce and small-batch production, the size of the intralogistics load unit of finished goods, semi-finished goods and raw materials is steadily shrinking. Totes are gradually replacing pallets as the primary handling and storage container. This shift has propelled tote-handling robotic systems to the forefront of automation order fulfillment centers. The order-fulfillment decisions of tote-handling robotic systems share a common order-tote-robot sequential decision-making nature. Existing studies primarily focus on decision mechanisms tailored to particular systems, making it difficult to generalize or transfer them to other contexts. We propose an Omni-scale Learning-based Sequential Decision Framework for Order Fulfillment of Tote-handling Robotic Systems (OLSF-TRS), a generalized and scalable sequential decision framework that combines structured combinatorial optimization with multi-agent reinforcement learning to coordinate order,tote, and robot decisions. On small-scale tote-handling robotic systems, OLSF-TRS achieves near-optimal performance with average optimality gaps below 3.5% across two distinct system configurations. In large-scale scenarios, OLSF-TRS consistently outperforms heuristic baselines across two different system types, reducing total tote movements by 8-12% and over 30% compared to SOTA rule-based approaches, while maintaining real-time responsiveness. These improvements translate into tangible operational benefits, including cost reduction, lower energy consumption, and enhanced throughput stability. The proposed framework delivers an efficient and unified order fulfillment decision-making framework for widely deployed tote-handling robotic systems,supporting high-quality order fulfillment in both e-commerce and industrial logistics sectors.

cs.RO↗
arXiv2026

SynerDiff: Synergetic Continuous Batching for Fast and Parallel Diffusion Model Inference

The expansion of Artificial Intelligence-generated content service requires diffusion model serving to simultaneously achieve high throughput and low task end-to-end (E2E) latency. However, existing continuous batching methods suffer from severe resource contention during UNet-VAE concurrency, leading to latency spikes. Furthermore, concurrent multi-task scheduling entails a trade-off between UNet throughput and VAE latency across varying scheduling strategies. To address these, we propose SynerDiff, an efficient continuous batching system built on intra-inter level synergy. At the intra-concurrency level, SynerDiff alleviates resource contention by pruning component-specific resource bottlenecks via VAE Chunking and Adaptive Skip-CFG. At the inter-concurrency level, leveraging components' differential sensitivity to scheduling granularities, a threshold-aware scheduler plans concurrent sequences and tunes intra-concurrency decisions to minimize VAE latency while maintaining UNet within high-throughput threshold. Additionally, a feedback controller dynamically adjusts this threshold based on queue loads to boost system capacity ceiling. Experimental results show that, SynerDiff improves throughput by 1.6$\times$ and decreases both average E2E and P99 tail latencies by up to 78.7\%, compared to benchmarks while guaranteeing high image fidelity.

cs.AI↗