arXiv subjects

Siyuan Zhang

Publications and source records attributed to Siyuan Zhang.

At least 19 recordsLinked to original sources

arXiv2026

Search for neutrinoless quadruple beta decay of $^{136}$Xe in PandaX-4T detector

The observation of neutrinoless quadruple beta decay (0$ν$4$β$) in the absence of neutrinoless double beta decay (0$ν$2$β$) has been argued to provide a strong indication that neutrinos are Dirac particles. We report a search for 0$ν$4$β$ decay of $^{136}\text{Xe}$ using a total $^{136}\text{Xe}$ exposure of 148.4 kg$\cdot$yr, collected during the commissioning and the first science runs of the PandaX-4T experiment. No significant excess of events over the background is observed. A lower limit on the 0$ν$4$β$ decay half-life of $^{136}\text{Xe}$ is set at 6.01 x $10^{24}$ yr at the 90% confidence level. This result establishes the most stringent constraint on this process in xenon, demonstrating the unique capability of the PandaX-4T detector in probing lepton number violation and shedding light on the fundamental nature of neutrinos.

nucl-ex↗

arXiv2026

Zonal dislocations in Laves phases: A coupled synchro-shear slip mechanism

Synchro-shear is the primary plastic deformation mechanism in Laves phases at elevated temperatures, mediated by synchro-Shockley partial dislocations-zonal dislocations that proceed via localized events such as kink-pair nucleation and propagation. Using atomistic simulations, we identified a novel slip mechanism in Laves phases, namely coupled synchro-shear slip, involving the synchronized glide of two synchro-Shockley partial dislocations on adjacent slip planes, leading to the formation of extrinsic stacking faults. High-resolution scanning transmission electron microscopy revealed the extended core structures consistent with coupled synchro-Shockley partial dislocations bounded by extrinsic stacking faults in C15 NbCr2 and their involvement in twinning. These results highlight the critical role of coupled synchro-shear slip in enabling phase transformations between Laves polytypes and in governing twinning behavior, providing new atomistic insight into the kinetic nature of plasticity in topologically close-packed intermetallic phases.

cond-mat.mtrl-sci↗

arXiv2026

ISO-RAG: Isoperimetric Noise Control for Retrieval-Augmented Generation

Retrieval-Augmented Generation (RAG) mitigates large language models (LLMs) hallucinations, yet conventional dense retrieval struggles with the complex reasoning paths of multi-hop question answering (QA). Graph-based RAG captures multi-step relationships but suffers from severe semantic drift and high online latency due to noisy global graph traversals. Thus, we propose ISO-RAG (ISOperimetric Retrieval-Augmented Generation), a geometry-aware RAG framework. By projecting the underlying knowledge graph into a hyperbolic Poincare ball to precompute node-wise isoperimetric profiles, ISO-RAG prunes spurious edges during retrieval, restricting the search space to a strictly localized subgraph. This topological purification regulates Personalized PageRank (PPR) diffusion driving the retrieval process, ensuring exact and low-latency convergence. Experiments on multi-hop QA benchmarks demonstrate that ISO-RAG outperforms state-of-the-art baselines by average absolute gains of 10.0% in retrieval recall and 4.3% in downstream exact match, achieving a superior accuracy-efficiency trade-off by fundamentally eliminating the latency bottleneck of global traversals. Our source code is available at https://github.com/ZaiizaiZHANG/ISO-RAG.

cs.AI↗

arXiv2026

When Safety Routing Breaks: Understanding Alignment Fragility under Benign Fine-Tuning

Benign fine-tuning severely weakens the safety alignment of large language models (LLMs), so we study why refusal behavior is so fragile. While prior work often attributes this failure to gradient conflict, we propose a fundamentally different Fisher-geometric explanation: safety Fisher is low-rank, and alignment makes the safety geometry flatter while preserving an output-routing pathway. After 100 benign fine-tuning examples, this pathway is selectively re-sharpened in output-side MLP modules, explaining the asymmetric fragility: safety can collapse to high attack success rates, while general utility degrades mildly. The routing view also explains why few safety examples can restore refusal behavior, indicating that internal safety-relevant representations are preserved. Finally, we show that LoRA and ASAM mitigate early collapse by suppressing output-side sharpness, but their protection weakens at larger fine-tuning scales. Overall, safety failure is best understood as a disruption of a low-rank output-routing mechanism

cs.CR↗

arXiv2026

Belief-Calibrated Optimization: An Explicit World Model for Agentic Optimization

The performance of an LLM agent depends on the scaffold around a frozen model. A common way to improve that scaffold is to use a coding agent as an optimizer: it reads current scores and traces and iteratively edits the source, producing a new candidate each round. Each edit is chosen according to a belief about how the environment will respond: what went wrong, and which change should help. That belief is typically implicit. It lives in the coding agent's reasoning on the current call, or remains latent in its parameters, rather than as something written down. Later calls therefore see scores and traces, but they do not use that belief. We introduce Belief-Calibrated Optimization (BCO), a method that writes that belief down as a persistent in-context document and continually revises that document as new candidates are evaluated. The resulting document is a world model: the current account of how the environment responds to edits. Added to an otherwise standard loop, BCO reaches a higher train passrate than a matched control that lacks only the world model, on five benchmarks spanning memory QA, tool-use QA, code-as-action app agents, and terminal agents. The gap remains on every held-out split, which is not used to select the candidate. After a target-model swap, in which the frozen model is replaced and the scaffold is not, the selected BCO scaffold leads on the tasks we test, except where context-window overruns leave it unfinished. An offline ablation then asks whether that gap comes from what the world model says. A fresh predictor given the accumulated document forecasts how the environment will respond more accurately than predictors given either no document or a same-form copy whose content has been falsified. The comparison indicates that the document carries reusable information in its content, not only in its form.

cs.AI↗

arXiv2026

EmoS: A Theory-Grounded Framework for Evaluating and Aligning Emotional Intelligence in Spoken Language Models

Despite significant advances in instruction-following and auditory comprehension, the evaluation of Emotional Intelligence (EI) in Spoken Language Models (SLMs) remains confined to rudimentary paralinguistic perception, lacking a systematic, theory-driven cognitive framework. We introduce EmoSBench, the first comprehensive EI evaluation benchmark for SLMs constructed upon the four-branch theoretical model, covering Perceiving, Understanding, Using, and Managing Emotion across ten sub-tasks. Preliminary assessments on EmoSBench reveal a substantial gap: even leading proprietary models like GPT-4o-Audio achieve only 52.6%, significantly trailing human baselines. To bridge this gap, we develop EmoS, a specialized evaluator model optimized via Supervised Fine-Tuning (SFT) and Group Relative Policy Optimization (GRPO). To facilitate its effective training, we curate EmoDialogue, a bilingual dataset providing necessary fine-grained supervision through response pairs with rigorously defined EI gradations. Concurrently, we introduce a reward mechanism integrating a Steep Exponential Accuracy Reward (SEAR) and a Rationale Fidelity Reward (RFR) to enforce precise ordinal scoring and valid reasoning. Experiments demonstrate that EmoS reaches 83.8% accuracy, approaching human-level performance. Furthermore, evaluations on authentic, unconstrained spoken interactions validate its robust real-world generalization, establishing a foundational framework for advancing emotionally intelligent dialogue systems.

cs.CL↗

arXiv2026

EBSD and Subtle Crystallographic Differences - A Study of Resolving Interlayer Spacings in Nb-Ni and Nb-Co mu-phases

In ordered intermetallics, slight variations in lattice site occupancy and specific interlayer spacings have been identified as the sources of significant changes in critical resolved shear stress and therefore how a given phase may affect alloy properties. So far, atom positions and lattice site occupancies have traditionally been characterised by high-resolution transmission electron microscopy (HR-TEM) and X-ray diffraction (XRD), which are methods that offer either local detail or high statistical significance but not both. Electron backscatter diffraction (EBSD), by contrast, provides high spatial resolution across large sample areas and therefore, has the potential to enable the local investigation of interlayer spacing and site lattice occupancy with improved statistical reliability. The objectives of the study are to benchmark EBSDs capability for resolving these subtle features and to correlate them with compositional and mechanical properties. In this case study, we therefore show that EBSD can resolve key crystallographic features of mu-phase intermetallics, specifically interlayer spacings. We combine pattern matching with large-scale dynamical simulations of template libraries guided by XRD based information on lattice parameters. For this, we generate structures that vary in the spacing between triple-layer and Kagome layer and in the site lattice occupancy of the 3a site. This approach successfully predicts the change of interlayer spacing between Kagome and triple layers in Nb-Co and Nb-Ni mu-phases, in good agreement with XRD and HR-TEM.

cond-mat.mtrl-sci↗

arXiv2026

EM3M: An Electron Micrograph Dataset for Microstructural Segmentation and Generation

Quantitative microstructural characterization is fundamental to materials science, and electron micrographs (EMs) provide indispensable high-resolution insights. However, progress in deep learning-based analysis of EMs has been hampered by the scarcity of large-scale, expert-annotated public datasets. To address this issue, we introduce EM3M, a large-scale and multimodal dataset for instance-level understanding of EMs. EM3M comprises 5,091 high-quality EMs, approximately 3 million instance segmentation annotations, and image-level textual descriptions with disentangled attributes. The dataset is constructed through a rigorous multi-stage curation and validation pipeline, with comprehensive statistical analyses to ensure reliability and reproducibility. Building upon these curated image-text pairs, we further provide a text-to-image diffusion model that serves as a controllable data augmentation engine, demonstrating that synthetic augmentation consistently improves downstream segmentation performance. To establish a systematic benchmark, we evaluate representative instance segmentation methods on EM3M. Our results reveal that conventional detection-based and query-based methods struggle with the extreme instance densities and textural complexities inherent in EMs. We additionally provide an optimized flow-based baseline to facilitate fair comparison and future research. EM3M {Dataset: https://huggingface.co/datasets/UniParser/EM3M}, the generative engine {Generation: https://huggingface.co/UniParser/EM3M-Gen}, and an online demo {Segmentation demo: https://www.bohrium.com/apps/uni-aims} are publicly available to support future research in automated materials analysis.

cs.CV↗

arXiv2026

Decentralized Stochastic Subgradient-type Methods with Communication Compression for Nonsmooth Nonconvex Optimization

In this paper, we consider the nonsmooth nonconvex decentralized optimization problem, where inter-agent communication is compressed. We propose a general framework that unifies various decentralized stochastic subgradient-type methods with unbiased compression and contractive compression with error compensation. By relating the consensus-error iterates and the averaged iterates to the trajectories of continuous-time differential inclusions, we establish global convergence for all methods encompassed by our framework when the objective functions are nonsmooth and lack Clarke regularity. Based on our framework, we further develop several compression-based methods, including decentralized stochastic subgradient methods utilizing sign-based regularization and gradient-tracking momentum. Preliminary numerical experiments empirically support our theoretical results and highlight the communication-accuracy trade-off of the newly developed methods.

math.OC↗

arXiv2026

Measurement of solar $pp$ neutrino flux with the new PandaX-4T data

We report a new measurement of the solar proton--proton ($pp$) neutrino flux via neutrino--electron elastic scattering using the PandaX-4T Run 2 data set collected between 2024 and 2026, corresponding to an exposure of 1.9 tonne$\cdot$yr. Before Run 2 data taking, the detector underwent a series of upgrades to improve its response and background conditions. Time variations of radioactive noble-gas impurities are constrained using the physics data themselves, complemented by measurements from the gas-assay system. The analysis introduced improvements in the data processing chain, detector response characterization, and background models. A blind spectral analysis was then performed on the electronic-recoil data across a wide energy range from 20 to 1000 keV. In combination with the Run 0 data published earlier, the fitted $pp$ flux is $(8.5 \pm 3.5)\times 10^{10}$ $\mathrm{cm^{-2}s^{-1}}$, consistent with the prediction of the Standard Solar Model. With a statistical significance of $2.2σ$ above background, this marks the first positive indication of solar $pp$ neutrino--electron scattering below an electronic-recoil energy of 165 keV.

hep-ex↗

arXiv2026

The Verification Horizon: No Silver Bullet for Coding Agent Rewards

A classical intuition holds that verifying a solution is easier than producing one. For today's coding agents, this intuition is being inverted: as foundation models develop stronger reasoning capabilities and engineering harnesses grow more sophisticated, generating complex candidate solutions is no longer difficult -- reliably verifying them has become the harder problem. Every verifier we can build is only a proxy for human intent, never the intent itself. This makes verification subject to a twofold difficulty: first, intent is underspecified by nature, making it inherently hard to faithfully check whether it has been fulfilled; second, during model training, optimization widens the gap between proxy and intent -- manifesting as reward hacking or signal saturation. To address this, we characterize the quality of verification signals along three dimensions -- scalability, faithfulness, and robustness -- and argue that achieving all three simultaneously is the central challenge. We further study four reward constructions: a test verifier for general coding tasks, a rubric verifier for frontend tasks, the user as verifier for real-world agent tasks, and an automated agent verifier for long-horizon tasks. Across different task types and policy capability levels, we conduct in-depth analysis and experiments on the core challenges of reward design and how to more effectively leverage reward signals. Experiments show that targeted verification design can effectively suppress reward hacking, improve task completion quality, and achieve significant gains across multiple internal and public benchmarks. These experiences collectively point to a core observation: no fixed reward function can remain effective as policy capability continues to grow; and verification must co-evolve with the generator.

cs.AI↗

arXiv2026

Material-Anisotropy-Driven Topological Optical Lattices on Thin-Film Lithium Niobate

Integrated structured-light sources usually obtain high-dimensional orbital angular momentum (OAM) states by encoding each channel into separate gratings, waveguides or metasurfaces, which ties modal capacity to structural complexity. Here we show that intrinsic material anisotropy can instead act as a built-in angular-momentum coupler. In an X-cut thin-film lithium niobate (TFLN) microring vortex emitter, the in-plane optical axis causes a circulating whispering-gallery mode to sample a periodically varying effective index, producing continuous azimuthal phase modulation. This modulation converts each resonance from a nominal single-charge emitter into a coherent topological sideband lattice with charges l=l_p+2n and Bessel-weighted amplitudes. Broadband measurements resolve a representative principal-charge series from l_p=-13 to +13, while additional devices with 100 and 200 GHz free spectral ranges (FSRs) show scalable resonance addressability. The emitted lattices are reproduced by a forward-calculated Fourier--Bessel model, supported by OAM projection measurements, and exhibit focusing into annular perfect-vortex fields and self-healing after obstruction. Waveguide-induced circular polarization further adds a vectorial spin--orbit channel. These results turn TFLN anisotropy from a material constraint into a compact mechanism for resonance-addressed high-dimensional structured-light generation.

physics.optics↗

arXiv2026

EChO-Agent: Evidence Chain Orchestration Agent for Audio Reasoning

While LALMs show promise on audio question answering, they fail to focus on question-relevant segments of audio and provide a clear, checkable reasoning process when dealing with complex audio reasoning. Reinforcement learning and tool-augmented prompting can help models better relate questions to audio but lack a reliable way to understand, integrate, and self-verify audio segments. To address this gap, we present EChO-Agent, a modular agent framework that reformulates complex audio QA as a planning, tool execution, evidence integration, and answer verification workflow. Experiments on MMAR benchmark show EChO-Agent improves both accuracy and rubric scores over baseline and ablation studies show evidence integration is the key factor.

eess.AS↗

arXiv2026

Topological antilaser

Coherent perfect absorption (CPA)-the time-reversed operation of lasing at threshold-relies on finely tuned interference and is intrinsically fragile to disorder and structural imperfections. Whether absorption can be endowed with topological protection, by analogy to topological lasing, has remained an open question. Here, we experimentally demonstrate a topological antilaser: the time-reversed counterpart of a topological laser, in which chiral edge modes of a photonic lattice enable perfect light absorption protected by topology. Using a nonreciprocal microwave network with low intrinsic loss, we show that the topological antilaser preserves near-unity absorption under strong disorder, and, unlike conventional antilasers, remains functional for arbitrary placements of dissipation and input ports, even when the lattice is strongly perturbed. This robustness arises from the disorder-immune propagation and stable spatial profile of the topological edge modes. Our results establish topologically protected absorption as the missing counterpart of topological lasing, opening new directions for studying robust energy dissipation, wave control, and coherent-absorption-based detection technologies.

physics.optics↗

arXiv2026

Probing scalar-neutrino and scalar-dark-matter interactions with PandaX-4T

Scalar-mediated interactions may exist among neutrinos, dark matter particles, or between the two. Double $β$-decay experiments provide a powerful tool to probe such exotic interactions. Using $^{136}$Xe double $β$-decay data from PandaX-4T, we perform the first direct spectral search in the energy range of 20 to 2800~keV, setting the most stringent limits to date on scalar-mediated neutrino self-interactions for mediator masses below 2~MeV$/c^2$. These results place significant constraints on models invoking such interactions to alleviate the Hubble Tension. Assuming the same scalar also mediates dark matter self-interactions, constraints on the dark matter-scalar interactions can be placed in conjunction with cosmological constraints.

hep-ex↗

arXiv2026

EXG: Self-Evolving Agents with Experience Graphs

Large language model (LLM)-based agents have demonstrated strong capabilities in complex reasoning and problem solving through multi-step interactions, yet most deployed agents remain behaviorally static, with knowledge acquired during execution rarely translating into systematic improvement over time. In response, a growing line of work on self-evolving agents explores how agents can improve through experience during deployment, but most existing approaches either rely on ad hoc reflection limited to single-task correction or adopt unstructured memory that accumulates fragmented experience with delayed usability. To address this limitation, we introduce EXG, an experience graph framework for self-evolving agents that explicitly organizes accumulated successes and failures into a structured, relational representation. EXG is the first experience graph designed for self-evolving agents, supporting both online, real-time graph growth during execution for immediate cross-task experience reuse, and offline reuse of a consolidated experience graph as an external memory module. This design also enables EXG to serve as a plug-and-play component for existing self-evolving agents, organizing prior experience into a unified experience graph and improving both solution quality and resource efficiency as deployment progresses. Extensive experiments across code generation and reasoning benchmarks show that EXG attains more favorable performance-efficiency trade-offs than reflection- and memory-based baselines in both online and offline evaluations. Our results suggest that structuring experience as a graph provides a principled foundation for scalable and transferable self-evolving agent behavior.

cs.AI↗

arXiv2026

Safety Alignment as Continual Learning: Mitigating the Alignment Tax via Orthogonal Gradient Projection

Safety post-training can improve the harmfulness and policy compliance of Large Language Models (LLMs), but it may also reduce general utility, a phenomenon often described as the \emph{alignment tax}. We study this trade-off through the lens of continual learning: sequential alignment stages expose the model to shifted data distributions and objectives, and their gradients may interfere with directions that support previously acquired general capabilities. This view does not claim that all alignment degradation has a single cause; rather, it provides a useful first-order mechanism for mitigating one important source of capability regression. We propose \textbf{O}rthogonal \textbf{G}radient \textbf{P}rojection for \textbf{S}afety \textbf{A}lignment (\textbf{OGPSA}), a lightweight update rule that estimates a low-rank reference subspace from gradients on a small set of general-capability data and removes from each safety gradient the component lying in this subspace. The resulting update is the steepest local safety-descent direction subject to first-order preservation constraints on the reference objectives. OGPSA is compatible with standard post-training pipelines and avoids large-scale replay, although it introduces periodic reference-gradient computation. Across Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and sequential SFT$\rightarrow$DPO settings, OGPSA improves the observed safety--utility trade-off over standard baselines. Under the sequential SFT$\rightarrow$DPO pipeline, the average performance gain increases from 33.98\% to 42.74\% on Qwen2.5-7B-Instruct and from 19.74\% to 32.98\% on Llama3.1-8B-Instruct. We have open sourced our code at https://github.com/SunGL001/OGPSA.

cs.LG↗

arXiv2026

Toward Exact Convergence in Byzantine-Robust Decentralized Learning: A Statistical Identification Approach

To defend against Byzantine attacks in decentralized learning, most existing methods rely on robust aggregation rules to mitigate the influence of malicious machines. However, these strategies inherently introduce bias, leading to inexact convergence with non-vanishing steady-state errors. In this paper, we propose a strategic shift from passive aggregation to active identification by introducing the Decentralized Rescaled Stochastic Gradient Descent with Byzantine Machine Identification (DRSGD-ByMI) framework. The core of our approach is an identification-based ``detect-then-optimize'' pipeline, where a p-value-free detection procedure is developed to accurately prune malicious nodes from the network. By leveraging sample-splitting score statistics, this identification mechanism achieves false discovery rate control without requiring restrictive distributional assumptions. We theoretically demonstrate that this precise identification allows the decentralized network to recover sufficient connectivity among the normal nodes, thereby enabling DRSGD-ByMI to match, even in the presence of Byzantine machines, the same order-optimal convergence rate as standard decentralized stochastic first-order methods. Numerical experiments validate our theoretical results and demonstrate the effectiveness of DRSGD-ByMI for decentralized robust learning problems.

stat.ME↗