Search arXiv⌕ Search

arXiv subjects

Yu Zhang

Publications and source records attributed to Yu Zhang.

At least 217 records · Page 12Linked to original sources
arXiv2026

Constraints and Consistency of Rotating Regular Black Hole Thermodynamics

The thermodynamics of regular black holes has long suffered from a self-consistency problem: the Hawking temperature derived from the first law disagrees with the geometrically defined one, signaling that the thermodynamic quantities obtained in previous studies are unreliable. To resolve this, we construct a rotating ``mother'' black hole whose parameters are initially independent, forming an extended thermodynamic phase space. Thermodynamic quantities are first derived in this unconstrained phase space, and only then is the regularity condition imposed. We find a fundamental asymmetry between the geometric and thermodynamic derivations: the geometric temperature is invariant under the order of constraint application, whereas the thermodynamic temperature requires the correct order of operations to yield a physically meaningful result. This establishes a self-consistent thermodynamic framework for rotating regular black holes, in which all thermodynamic quantities are guaranteed to be physically correct, laying a rigorous foundation for investigating phase transitions and thermodynamic stability.

gr-qc↗
arXiv2026

Proton High-Order Cumulants in Au+Au Collisions at High Baryon Density from JAM with a Centrality-Independent Framework

The event-by-event higher-order cumulants of conserved quantities such as net-baryon, net-electric charge, and net-strangeness in heavy-ion collisions have been extensively utilized in experimental searches for the QCD critical point, notably in the RHIC-STAR experiment. In this study, we conduct a systematic analysis of higher-order cumulants of proton number distributions in Au+Au collisions at center-of-mass energies of $\sqrt{s_{\rm NN}} = 3.2$, $3.5$, $3.9$, and $4.5$ GeV using the JAM model. We calculate cumulants, factorial cumulants, and their ratios using a novel method, Centrality-Independent Genuine Cumulant Analysis fRamework (CIGAR), which effectively eliminates initial volume fluctuations. We comprehensively compare the CIGAR method with the traditional Centrality Bin Width Correction (CBWC) method. In addition, the effect of spectators on cumulant is systematically investigated. Our results provide a dynamic non-critical baseline in the high-baryon-density regime which is crucial for QCD critical point searches in heavy-ion collisions.

physics.data-an↗
arXiv2026

Towards Streaming Synchronized Spatial Audio Generation via Autoregressive Diffusion Transformer

Real-time and accurate spatial audio generation is pivotal for delivering an immersive experience. However, existing spatial audio synthesis technologies are often encumbered by a tradeoff between generation quality and high inference latency, as well as difficulty in capturing precise spatial information from multimodal inputs. To address these challenges, we propose SwanSphere, a unified streaming framework for high-fidelity spatial audio generation from panoramic videos and text prompts. SwanSphere mainly makes the following contributions: 1) We introduce a causal autoregressive diffusion transformer architecture that enables streaming high-quality spatial audio generation. 2) We design a Spatial Video-Audio Contrastive (SVAC) learning strategy to align the video encoder with the acoustic domain, and further employ a multi-objective online direct preference optimization (ODPO) scheme, resulting in strong spatial perception and robust multimodal spatial audio synthesis. 3) To alleviate the current scarcity of spatial audio datasets, we also develop an automated annotation pipeline for generating detailed spatial captions. Experimental results demonstrate that SwanSphere achieves superior performance in both video-to-spatial and text-to-spatial audio generation tasks. Demos can be found at: https://swanaigc.github.io.

eess.AS↗
arXiv2026

SwanVoice: Expressive Long-Form Zero-Shot Speech Synthesis for Both Monologue and Dialogue

Zero-shot text-to-speech (TTS) has improved substantially for single-speaker synthesis, yet expressive long-form multi-speaker dialogue remains difficult. A common workaround is to synthesize each turn with a monologue TTS model and stitch the outputs together. This adds inference cost and often breaks acoustic consistency, conversational coherence, and affective continuity across turns. Recent dialogue TTS systems have begun to address this setting, but they still struggle to keep expressive coherence, controllable speaker switching, and monologue quality at the same time. We present SwanData-Speech and SwanVoice. SwanData-Speech builds monologue and dialogue corpora from in-the-wild audio, using Swan Forced Aligner for pause-aware word-level alignment and RobustMegaTTS3 for pronunciation-hard cases. Built on these data, SwanVoice is a zero-shot TTS model for 1--4 speakers, combining a 25 Hz VAE, raw-text conditioning with pause-aware symbols and pinyin substitution, and a flow-matching DiT with speaker-turn conditioning. Training starts from monologue speech, moves through mixed and real dialogue data, and then uses DiffusionNFT post-training with phone-level and speaker-similarity rewards. On SwanBench-Speech, SwanVoice obtains higher richness and hierarchy scores than all evaluated open-source baselines in both monologue and dialogue settings, while content accuracy remains the main limitation. Audio demos are available at https://swanaigc.github.io//#swanvoice.

eess.AS↗
arXiv2026

Masking Stale Observations Helps Search Agents -- Until It Doesn't: A Regime Map and Its Mechanism

Long-horizon search agents accumulate large amounts of retrieved content across many tool calls, making context-budget efficiency increasingly important. A minimal intervention is to mask stale observations from the context as the trajectory progresses, but it remains unclear when this form of context management helps and why. We study observation masking through a systematic sweep over various agent backbones (4B to 284B parameters) and three retrievers on offline and live-web agentic search benchmarks. We find that the accuracy gain from masking follows an asymmetric inverted-U shape when plotted against the model's accuracy without context management: a plateau under weak retrievers, a peak when a strong retriever meets a mid-capacity model, and a sharp collapse when the model is saturated. This pattern reflects the interaction between retriever recall and the model's implicit filtering capacity, rather than either factor in isolation. Mechanistically, masking implements a token-for-turn trade-off: it removes observations the model has largely stopped attending to and pages the agent rarely re-opens. The added turns help when they convert failures into successes, but they fail when masking removes evidence the model would otherwise have used. We therefore reframe context management as a regime-dependent intervention and provide a holistic perspective for analyzing context use in agentic deep search. We release our scaffold and trajectories here (https://github.com/i-DeepSearch/observation-masking) to support future research.

cs.CL↗
arXiv2026

MIMO-OTFS-Based Semantic Communication for High-Mobility Scenarios

In high-mobility scenarios with time-frequency doubly-selective channels, existing semantic communication systems suffer significant performance degradation. To address this issue, we propose a semantic communication framework that synergistically integrates multiple-input multiple-output orthogonal time frequency space (MIMO-OTFS) with semantic-aware sub-channel allocation. First, an entropy module is employed to evaluate importance of different semantic features, and the Kendall correlation coefficient is used to quantify the alignment between semantic importance and sub-channel conditions. Subsequently, joint optimization of the encoder and decoder is achieved through a comprehensive loss function that balances image classification accuracy, reconstruction quality, and sub-channel matching degree. Experimental results confirm the superior reconstruction quality of our proposed framework compared to conventional semantic communication systems based on orthogonal frequency division multiplexing in high-mobility channel environment.

eess.SP↗
arXiv2026

User-Centric Clustering for uRLLC in Cell-Free RAN via Extreme Value Theory

Ultra-reliable low-latency communication (uRLLC) is a pivotal enabler for B5G/6G networks, yet it faces severe challenges from rare but critical extreme events, which are characterized by heavy tails in the delay distribution. While the cell-free radio access network (CF-RAN) architecture offers essential spatial diversity to combat these uncertainties, conventional user-centric clustering designs typically focus on average metrics, thereby inadequately addressing such tail behaviors. We propose a novel, tail-risk-aware, user-centric clustering framework operating within the finite blocklength (FBL) regime. Our approach employs extreme value theory (EVT), specifically the peaks-over-threshold (POT) model, to accurately quantify the probability of queue latency violations. This framework is applied to formulate an energy efficiency (EE) maximization problem under strict tail latency constraints. The problem is solved via an efficient online algorithm that integrates Lyapunov optimization with successive convex approximation (SCA). Simulation results demonstrate that the proposed scheme, through its dynamic adaptation of cluster formation to mitigate tail risks, achieves a superior reliability-efficiency trade-off and leads to a significant suppression of extreme latency events.

cs.IT↗
arXiv2026

Column-Wise Analog Processing for Hybrid Precoding in Millimeter Wave Downlink Multi-User Massive MIMO Systems

In millimeter wave (mmWave) massive MIMO systems, existing alternating minimization (AltMin) based hybrid digital-analog precoding algorithms can achieve near-optimal spectral efficiency (SE) of the fully-digital precoding. However, this kind of AltMin algorithms require several hundreds of iterations to optimize the analog precoder, thus increasing the complexity. This paper focuses on reducing complexity of the analog precoding and proposes a column-wise analog precoding (CWAP) algorithm. The main idea is to seek closed-form solution of the analog precoder, through which the analog precoder can be easily computed in one step instead of iterations, thus reducing the complexity. Specifically, by assuming that perfect digital precoder is deployed at the base station (BS) to eliminate interferences among users, we simplify the expression of achievable SE for each user. Subsequently, the simplified SE is further converted to the sum of a series of sub-rates, each of which is related to the corresponding column of the analog precoder. The optimization problem of maximizing the SE is then transformed into a series of sub-problems of maximizing each sub-rate. Upon solving each sub-problem, closed-form solution of each column of the analog precoder can be directly obtained without iterations, resulting in reduced complexity. Simulation results demonstrate that (a) when the number of RF chains equals the number of data streams, the proposed scheme can achieve approximately same sum-rate as the AltMin algorithms; (b) when the number of RF chains is larger than the number of data streams, the proposed scheme can achieve higher sum-rate than the AltMin algorithms; (c) the proposed scheme has lower complexity than the AltMin algorithms (almost one order of magnitude reduction in some cases).

eess.SP↗
arXiv2026

NaRA: Noise-Aware LoRA for Parameter-Efficient Fine-Tuning of Diffusion LLMs

Diffusion Large Language Models (dLLMs) have emerged as a promising non-autoregressive generative paradigm. Given the prohibitive computational cost of full fine-tuning, Parameter-Efficient Fine-Tuning (PEFT) has become the standard approach. However, existing PEFT methods (e.g., LoRA), originally tailored for autoregressive models, rely on static parameters that are agnostic to the noise level. Consequently, they ignore the intrinsic dynamics of the diffusion process, where input distributions and generation difficulty shift significantly along the denoising trajectory, rendering them suboptimal for dLLMs. To address this, we propose Noise-aware Low-Rank Adaptation (NaRA), which introduces a low-rank core matrix generated by a lightweight, globally shared hypernetwork conditioned on the noise level. This design enables the update matrices to vary continuously along the diffusion process while keeping parameter and latency overhead negligible. We provide a theoretical justification for the proposed NaRA framework and empirically demonstrate consistent improvements over noise-agnostic baselines across commonsense reasoning, mathematical reasoning, and code generation benchmarks. Our code is available at https://github.com/generaldi/NaRA.

cs.AI↗
arXiv2026

Programmable Dissipation via Partial Quantum Error Correction

Noise is typically treated as the adversary of quantum information processing. For open quantum dynamics, however, dissipation is part of the target physics, creating a tension with fault-tolerant architectures designed to suppress decoherence. Here we show that logical noise can instead be turned into a calibrated resource. We treat the error-correction cycle as a programmable primitive: one fault-tolerant round induces a logical completely positive trace-preserving map, and decoder/recovery randomization generates a controllable family of logical channels whose convex mixtures realize Kraus-channel mixing. This enables direct compilation of target dissipators into effective logical dynamics without explicit ancilla qubits for encoding the bath degree of freedoms. We derive an accuracy criterion for multi-step simulation in which the code distance is chosen so that uncontrolled logical errors remain a small fraction of the intended dissipation per step, rather than being driven below an arbitrarily small closed-system tolerance. Partial quantum error correction thus repurposes fault-tolerant structure to sculpt dissipation, offering a resource-efficient route to quantum simulation of open quantum systems.

quant-ph↗
arXiv2026

XTransfer: Modality-Agnostic Few-Shot Model Transfer for Human Sensing at the Edge

Deep learning for human sensing on edge systems presents significant potential for smart applications. However, its training and development are hindered by the limited availability of sensor data and resource constraints of edge systems. While transferring pre-trained models to different sensing applications is promising, existing methods often require extensive sensor data and computational resources, resulting in high costs and limited transferability. In this paper, we propose XTransfer, a first-of-its-kind method enabling modality-agnostic, few-shot model transfer with resource-efficient design. XTransfer flexibly uses pre-trained models and transfers knowledge across different modalities by (i) model repairing that safely mitigates modality shift by adapting pre-trained layers with only few sensor data, and (ii) layer recombining that efficiently searches and recombines layers of interest from source models in a layer-wise manner to restructure models. We benchmark various baselines across diverse human sensing datasets spanning different modalities. The results show that XTransfer achieves state-of-the-art performance while significantly reducing the costs of sensor data collection, model training, and edge deployment.

cs.CV↗
arXiv2026

Large-Scale AI and Foundation Models for Neuroscience: A Comprehensive Review

The development of large-scale artificial intelligence (AI) models is influencing neuroscience research by enabling end-to-end learning from raw brain signals and neural data. In this paper, we review applications of large-scale AI models across four major neuroscience domains: neuroimaging and data processing, brain-computer interfaces and neural decoding, clinical decision support and translational frameworks, and disease-specific applications across neurological and psychiatric disorders. These models show potential to address major computational neuroscience challenges, including multimodal neural data integration, spatiotemporal pattern interpretation, and the development of translational frameworks for clinical research. Moreover, the interaction between neuroscience and AI has become increasingly reciprocal, as biologically informed architectural constraints are now incorporated to develop more interpretable and computationally efficient models. This review highlights both the promise of such technologies and critical implementation considerations, with particular emphasis on rigorous evaluation frameworks, effective integration of domain knowledge, prospective clinical validation, and comprehensive ethical guidelines. Finally, a systematic listing of critical neuroscience datasets used to develop and evaluate large-scale AI models across diverse research applications is provided.

cs.AI↗
arXiv2026

SPARD: Defending Harmful Fine-Tuning Attack via Safety Projection with Relevance-Diversity Data Selection

Fine-tuning large language models often undermines their safety alignment, a problem further amplified by harmful fine-tuning attacks in which adversarial data removes safeguards and induces unsafe behaviors. We propose SPARD, a defense framework that integrates Safety-Projected Alternating optimization with Relevance-Diversity aware data selection. SPARD employs SPAG, which optimizes alternatively between utility updates and explicit safety projections with a set of safe data to enforce safety constraints. To curate safe data, we introduce a Relevance-Diversity Determinantal Point Process to select compact safe data, balancing task relevance and safety coverage. Experiments on GSM8K and OpenBookQA under four harmful fine-tuning attacks demonstrate that SPARD consistently achieves the lowest average attack success rates, substantially outperforming state-of-the-art defense methods, while maintaining high task accuracy. Code is available at https://github.com/shuhao02/SPARD.

cs.LG↗
arXiv2026

Comprehensive Benchmarking of Long-Form Speech Generation in Diverse Scenarios

Recent advances in speech generation have enabled high-fidelity synthesis, yet systematic evaluation of models under long-context conditions remains largely underexplored. A comprehensive evaluation benchmark for long-form speech is indispensable for two reasons: 1) existing test scenarios are often confined to limited domains, creating a significant gap with the diverse downstream applications; 2) existing metrics overlook critical long-text factors such as consistency and coherence, failing to generalize reliably. To this end, we propose Swanbench-Speech, a comprehensive benchmark that decomposes long-form speech quality into specific, disentangled dimensions. SwanBench-Speech has three key properties. 1) Rich speech scenarios: Focusing on long-form speech generation and dialog generation, SwanBench-Speech covers acoustics, semantics, and expressiveness challenges, and consists of 1,101 samples spanning 17 common speech scenarios; 2) Comprehensive evaluation dimensions: Along the acoustics, semantics, and expressiveness axes, SwanBench-Speech defines an automated evaluation protocol with seven metrics to provide a comprehensive, accurate, and standardized assessment; 3) Valuable Insights: Through extensive experiments, we reveal that current models still struggle in highly expressive scenarios and exhibit a notable gap in consistency and hierarchy compared to real recordings.

eess.AS↗
arXiv2026

MATT-CTR: Unleashing a Model-Agnostic Test-Time Paradigm for CTR Prediction with Confidence-Guided Inference Paths

Recently, a growing body of research has focused on either optimizing CTR model architectures to better model feature interactions or refining training objectives to aid parameter learning, thereby achieving better predictive performance. However, previous efforts have primarily focused on the training phase, largely neglecting opportunities for optimization during the inference phase. Infrequently occurring feature combinations, in particular, can degrade prediction performance, leading to unreliable or low-confidence outputs. To unlock the predictive potential of trained CTR models, we propose a Model-Agnostic Test-Time paradigm (MATT), which leverages the confidence scores of feature combinations to guide the generation of multiple inference paths, thereby mitigating the influence of low-confidence features on the final prediction. Specifically, to quantify the confidence of feature combinations, we introduce a hierarchical probabilistic hashing method to estimate the occurrence frequencies of feature combinations at various orders, which serve as their corresponding confidence scores. Then, using the confidence scores as sampling probabilities, we generate multiple instance-specific inference paths through iterative sampling and subsequently aggregate the prediction scores from multiple paths to conduct robust predictions. Finally, extensive offline experiments and online A/B tests strongly validate the compatibility and effectiveness of MATT across existing CTR models.

cs.LG↗
arXiv2026

Drive-P2D: A Progressive Perception-to-Decision Benchmark for VLMs in Autonomous Driving

Autonomous driving requires reliable perception and safe decision-making in complex scenarios. Recent vision-language models (VLMs) demonstrate reasoning and generalization abilities, opening new possibilities for autonomous driving; however, existing benchmarks often evaluate perception and decision-making separately, limit failure analysis with choice-only formats, or introduce evaluation bias through LLM-scored long-form outputs. To address these issues, we present Drive-P2D, a progressive perception-to-decision benchmark with 6,650 questions across Object, Scene, and Decision levels. Drive-P2D adopts a separated reasoning-and-answer protocol: final answers are scored objectively, while reasoning is analyzed to identify error modes exposed along the progressive perception-to-decision chain. We evaluate mainstream VLMs across all and high-risk scenarios, and further characterize the perception-to-decision capability boundary through correlation analysis and similar-scene robustness testing. Reasoning further exposes failure modes such as logical reasoning errors and semantic feature omissions, and we train a lightweight analyzer model to automate large-scale error-mode annotation of reasoning. Together, these designs provide practical insights for building safer and more reliable VLMs for real-world autonomous driving.

cs.AI↗
arXiv2026

Verilog-Evolve: Feedback-Driven and Skill-Evolving Verilog Generation

Large language models (LLMs) have improved Verilog generation from natural-language specifications, but most pipelines still treat generation as isolated sampling followed by functional checking. This is insufficient for practical RTL design, where useful Verilog must be correct, synthesizable, timing-conscious, and friendly to downstream hardware objectives. We present Verilog-Evolve, a feedback-driven framework for versioned Verilog refinement and cross-session skill evolution. For each task, Verilog-Evolve generates diverse minor candidates, evaluates them with executable feedback from functional simulation, Yosys synthesis, ABC timing proxy, and optional GEMM metrics, then promotes the best candidate into a major version under configurable scoring. To improve across tasks, the system maintains modular skill guidance, retrieves skills according to task and feedback context, and evolves candidate skills from logged histories through create/improve/skip decisions and verifier reports. Experiments on VerilogEval and mixed-precision GEMM tasks show that Verilog-Evolve improves final functional success and promotion stability while producing more downstream-friendly RTL under open-source synthesis, timing-proxy, and netlist-level GEMM objectives. Validation-gated skill evolution further improves GEMM downstream quality and achieves the best downstream score and GEMM held-out pass rate among the evaluated skill modes.

cs.CL↗
arXiv2026

Therm-FM: Foundation Model is ALL YOU NEED for 3D-ICs Thermal Simulation

Data-driven thermal predictors for 3D-ICs are often trained from scratch for each chip design using many high-fidelity finite-element simulations, leading to high data-generation cost and costly cross-design reuse. We propose Therm-FM, a neural operator framework that adapts a pretrained partial differential equation (PDE) foundation model to steady-state and transient 3D-IC thermal simulation. The motivation is that steady-state and transient chip-level heat conduction respectively share elliptic and parabolic operator structures with diffusion-type PDEs, allowing pretrained diffusion priors to provide an effective initialization for thermal-field prediction under heterogeneous materials, dense TSV/microbump interconnects, and package-level boundary conditions. To further reduce data-generation cost, Therm-FM incorporates a thermal-equivalent multi-fidelity training strategy that uses low-cost approximate simulations for thermal-domain adaptation and limited high-fidelity samples for calibration. Experiments on public HotSpot benchmarks and industrial 3D-IC package benchmarks show that Therm-FM achieves up to a 10.6x reduction in mean error and surpasses prior best accuracy with less than 20% of the training data. In cross-chip adaptation, it matches or surpasses full-data baselines in several metrics using only 10--30 target samples. We release datasets, source code, and pretrained models at https://github.com/haiyangxin/Therm-FM.

cs.CE↗