Searcharxiv⌕ Search

arXiv subjects

Zheyu Wang

Publications and source records attributed to Zheyu Wang.

At least 19 recordsLinked to original sources
arXiv2026

Thickness-driven crossover from conventional to chiral nonreciprocal superconductivity in kagome metal CsV3Sb5

Superconductivity and its potential applications are governed by the symmetry of the superconducting order parameter. In the kagome metal CsV3Sb5, most bulk studies indicate conventional s-wave pairing. However, ultrathin flakes exhibit nonreciprocal transport, in particular a zero-field superconducting diode effect, which requires broken inversion and time-reversal symmetries. Here, using thickness dependent transport measurements, we observe the emergence of non-reciprocal second-harmonic magnetotransport signals and a zero-field superconducting diode effect, accompanied by a pronounced reduction of the out-of-plane coherence length with decreasing thickness. Upper critical field measurements further reveal a dimensional crossover from three-dimensional superconductivity in bulk to two-dimensional superconductivity in thin flakes. These findings indicate a thickness-induced chiral superconducting phase that breaks both inversion and time-reversal symmetries in the two-dimensional limit. Our work not only clarifies long-standing controversies regarding the pairing symmetry in CsV3Sb5, but also establishes thin-flake kagome superconductors as a versatile platform for engineering nonreciprocal quantum devices and exploring emergent topological phases.

cond-mat.supr-con↗
arXiv2026

Anatomy-Slot: Unsupervised Anatomical Factorization for Homologous Bilateral Reasoning in Retinal Diagnosis

Retinal diagnosis is inherently bilateral: clinicians compare homologous structures across eyes (e.g., optic disc asymmetry), yet most deep models operate on monocular representations. We investigate whether explicit structural correspondence improves diagnosis, and propose Anatomy-Slot to operationalize this hypothesis. Anatomy-Slot introduces an unsupervised anatomical bottleneck by decomposing patch tokens into a set of emergent, structurally-coherent slots that correspond to anatomical regions, then aligning these slots across eyes via bidirectional cross-attention. On ODIR-5K with $n=10$ seeds, the method improves AUC by $4.2$ points over a matched ViT-L baseline (95% CIs; Wilcoxon signed-rank test, $W=0$, $p=0.002$). Pairing disruption and stress testing under Gaussian noise provide controlled tests of correspondence dependence and robustness under corruption. We further report quantitative optic disc grounding on REFUGE and cross-attention localization analysis. Beyond the reported gains, these results indicate that object-centric anatomical correspondence offers a principled path toward interpretable diagnostic systems aligned with clinical bilateral comparison.

cs.CV↗
arXiv2026

Anomalous Low-temperature Magnetotransport in Kagome Metal CsCr$_3$Sb$_5$ under Pressure

As a unique kagome superconductor displaying clear signatures of strong electronic correlations, CsCr$_3$Sb$_5$ has drawn much attention. Its rich temperature-pressure phase diagram features intertwined orders including pressure-induced superconductivity and two density-wave-like phases, making it an outstanding platform to explore the complex coexistence and competition of multiple quantum orders. At around 30 K, which we designate as $T_3$, a possible anomaly manifesting as a hump in the resistivity has been observed, yet its nature remains largely unexplored due to limited supporting evidence from other probes. Here, we conducted systematic magnetotransport experiments under hydrostatic pressure to investigate the nature of this anomaly. Our results reveal an abundance of intriguing magnetotransport signatures below $T_3$, including a non-trivial temperature dependence of the Hall coefficient, multi-band characteristics, and pressure-enhanced anomalous-Hall-like effect. These signatures bear resemblance to those observed in the charge-density-wave state in the sister compound CsV$_3$Sb$_5$. These findings suggest the possibility of an additional, exotic electronic order in CsCr$_3$Sb$_5$, calling for further detailed investigations.

cond-mat.str-el↗
arXiv2026

BTC-LLM: Efficient Sub-1-Bit LLM Quantization via Learnable Transformation and Binary Codebook

Binary quantization represents the most extreme form of compression, reducing weights to +/-1 for maximal memory and computational efficiency. While recent sparsity-aware binarization achieves sub-1-bit compression via weight pruning, it faces critical challenges: performance degradation, mask-management overhead, and limited hardware compatibility. In this paper, we present BTC-LLM, a novel sub-1-bit LLM quantization framework that leverages binary pattern clustering and weight transformation to overcome these limitations. Our approach incorporates two key innovations: (1) a Binary Codebook that clusters recurring vectors into compact indices using custom distance metrics and sign-based updates; (2) a Learnable Transformation that reduces outliers and promotes shared sign patterns among binary weights. This eliminates sparse masks, enabling efficient inference on standard hardware. Extensive evaluations across LLaMA, Qwen, and FBI-LLM families demonstrate that BTC-LLM achieves state-of-the-art results in extreme compression (1.11-0.7 bits). Notably, BTC-LLM compressed to 0.8 bits on LLaMA-2-13B maintains high performance, with only a 3.1 percent accuracy drop in zero-shot benchmarks, while delivering a 1.6x speedup over FP16.

cs.LG↗
arXiv2026

BABE: Biology Arena BEnchmark

The rapid evolution of large language models (LLMs) has expanded their capabilities from basic dialogue to advanced scientific reasoning. However, existing benchmarks in biology often fail to assess a critical skill required of researchers: the ability to integrate experimental results with contextual knowledge to derive meaningful conclusions. To address this gap, we introduce BABE(Biology Arena BEnchmark), a comprehensive benchmark designed to evaluate the experimental reasoning capabilities of biological AI systems. BABE is uniquely constructed from peer-reviewed research papers and real-world biological studies, ensuring that tasks reflect the complexity and interdisciplinary nature of actual scientific inquiry. BABE challenges models to perform causal reasoning and cross-scale inference. Our benchmark provides a robust framework for assessing how well AI systems can reason like practicing scientists, offering a more authentic measure of their potential to contribute to biological research.

cs.AI↗
arXiv2026

OMG-Agent: Toward Robust Missing Modality Generation with Decoupled Coarse-to-Fine Agentic Workflows

Data incompleteness severely impedes the reliability of multimodal systems. Existing reconstruction methods face distinct bottlenecks: conventional parametric/generative models are prone to hallucinations due to over-reliance on internal memory, while retrieval-augmented frameworks struggle with retrieval rigidity. Critically, these end-to-end architectures are fundamentally constrained by Semantic-Detail Entanglement -- a structural conflict between logical reasoning and signal synthesis that compromises fidelity. In this paper, we present \textbf{\underline{O}}mni-\textbf{\underline{M}}odality \textbf{\underline{G}}eneration Agent (\textbf{OMG-Agent}), a novel framework that shifts the paradigm from static mapping to a dynamic coarse-to-fine Agentic Workflow. By mimicking a \textit{deliberate-then-act} cognitive process, OMG-Agent explicitly decouples the task into three synergistic stages: (1) an MLLM-driven Semantic Planner that resolves input ambiguity via Progressive Contextual Reasoning, creating a deterministic structured semantic plan; (2) a non-parametric Evidence Retriever that grounds abstract semantics in external knowledge; and (3) a Retrieval-Injected Executor that utilizes retrieved evidence as flexible feature prompts to overcome rigidity and synthesize high-fidelity details. Extensive experiments on multiple benchmarks demonstrate that OMG-Agent consistently surpasses state-of-the-art methods, maintaining robustness under extreme missingness, e.g., a $2.6$-point gain on CMU-MOSI at $70$\% missing rates.

cs.AI↗
arXiv2026

Retrieval-Infused Reasoning Sandbox: A Benchmark for Decoupling Retrieval and Reasoning Capabilities

Despite strong performance on existing benchmarks, it remains unclear whether large language models can reason over genuinely novel scientific information. Most evaluations score end-to-end RAG pipelines, where reasoning is confounded with retrieval and toolchain choices, and the signal is further contaminated by parametric memorization and open-web volatility. We introduce DeR2, a controlled deep-research sandbox that isolates document-grounded reasoning while preserving core difficulties of deep search: multi-step synthesis, denoising, and evidence-based conclusion making. DeR2 decouples evidence access from reasoning via four regimes--Instruction-only, Concepts (gold concepts without documents), Related-only (only relevant documents), and Full-set (relevant documents plus topically related distractors)--yielding interpretable regime gaps that operationalize retrieval loss vs. reasoning loss and enable fine-grained error attribution. To prevent parametric leakage, we apply a two-phase validation that requires parametric failure without evidence while ensuring oracle-concept solvability. To ensure reproducibility, each instance provides a frozen document library (drawn from 2023-2025 theoretical papers) with expert-annotated concepts and validated rationales. Experiments across a diverse set of state-of-the-art foundation models reveal substantial variation and significant headroom: some models exhibit mode-switch fragility, performing worse with the Full-set than with Instruction-only, while others show structural concept misuse, correctly naming concepts but failing to execute them as procedures.

cs.AI↗
arXiv2026

HeartMuLa: A Family of Open Sourced Music Foundation Models

We present a family of open-source Music Foundation Models designed to advance large-scale music understanding and generation across diverse tasks and modalities. Our framework consists of four major components: (1) HeartCLAP, an audio-text alignment model; (2) HeartTranscriptor, a robust lyric recognition model optimized for real-world music scenarios; and (3) HeartCodec, a low-frame-rate (12.5 Hz) yet high-fidelity music codec tokenizer that captures long-range musical structure while preserving fine-grained acoustic details and enabling efficient autoregressive modeling; (4) HeartMuLa, an LLM-based song generation model capable of synthesizing high-fidelity music under rich, user-controllable conditions (e.g., textual style descriptions, lyrics, and reference audio). In addition, it provides two specialized modes: (i) fine-grained musical attribute control, which allows users to specify the style of different song sections (e.g., intro, verse, chorus) using natural language prompts; and (ii) short, engaging music generation, which is suitable as background music for short videos. Lastly, HeartMuLa improves significantly when scaled to 7B parameters. For the first time, we show that a Suno-level, commercial-grade system can be reproduced using academic-scale data and GPU resources. We expect these foundation models to serve as strong baselines for future research and to facilitate practical applications in multimodal content production.

cs.SD↗
arXiv2025

Giant critical current peak induced by pressure in kagome superconductor RbV$_{3}$Sb$_{5}$

Superconductivity can coexist or compete with other orders such as magnetism or density waves. Optimizing superconductivity requires identifying competing orders that may disrupt Cooper pair coherence. Here, we use the self-field critical current ($I_{\rm c,sf}$) to probe pressure-tuned superconductivity in the kagome superconductor RbV$_3$Sb$_5$. As pressure destabilizes the charge-density wave (CDW) state, $I_{\rm c,sf}$ drastically enhances, peaking near the critical pressure where the CDW state is completely suppressed at zero temperature. Surprisingly, a weaker $I_{\rm c,sf}$ peak emerges within the CDW phase. Near the pressure of the weaker peak, the superconducting phase transition temperature shifts from an increasing trend with pressure to a near plateau. Our analysis suggests the possibility of a sudden change in the CDW pattern or a Lifshitz transition, highlighting the need for microscopic examinations of the CDW state for understanding the pressure evolution of superconductivity in RbV$_3$Sb$_5$.

cond-mat.supr-con↗
arXiv2025

Recitation over Reasoning: How Cutting-Edge Language Models Can Fail on Elementary School-Level Reasoning Problems?

The rapid escalation from elementary school-level to frontier problems of the difficulty for LLM benchmarks in recent years have weaved a miracle for researchers that we are only inches away from surpassing human intelligence. However, is the LLMs' remarkable reasoning ability indeed comes from true intelligence by human standards, or are they simply reciting solutions witnessed during training at an Internet level? To study this problem, we propose RoR-Bench, a novel, multi-modal benchmark for detecting LLM's recitation behavior when asked simple reasoning problems but with conditions subtly shifted, and conduct empirical analysis on our benchmark. Surprisingly, we found existing cutting-edge LLMs unanimously exhibits extremely severe recitation behavior; by changing one phrase in the condition, top models such as OpenAI-o1 and DeepSeek-R1 can suffer 60 percent performance loss on elementary school-level arithmetic and reasoning problems. Such findings are a wake-up call to the LLM community that compels us to re-evaluate the true intelligence level of cutting-edge LLMs.

cs.AI↗
arXiv2025

Point-contact Andreev reflection spectroscopy of layered superconductors with device-integrated diamond anvil cells

Superconductors that can be mechanically exfoliated are an interesting platform for exploring superconducting properties tuned by layer thickness. These layered superconductors are also expected to exhibit sensitivity to applied pressure. While pressure has been demonstrated to be an effective way of tuning bulk superconductors, analogous studies on superconducting thin flakes have been limited due to technical challenges. In particular, spectroscopic measurements under pressure remain insufficiently explored. In this work, we functionalized the diamond anvil cell technique for point-contact Andreev reflection spectroscopy (PCAR) measurement on thin-flake materials under pressure, offering the opportunity to obtain spectroscopic information on superconductivity. To validate the feasibility of this method, we have conducted PCAR measurements on iron-selenide thin flakes to extract temperature-dependent superconducting gap values under ambient and high pressure. Combine with the proven magnetotransport capability, our method provides a conceptually simple tool for a detailed examination of thin-flake superconductors under pressure.

cond-mat.supr-con↗
arXiv2025

Superconducting critical temperature and dimensionality tuning of RbV$_3$Sb$_5$ via biaxial strain

Kagome metal AV$_3$Sb$_5$ (A=K, Rb, Cs) has emerged as an intriguing platform for exploring the interplay between superconductivity and other quantum states. Among the three compounds, RbV$_3$Sb$_5$ has a notably lower superconducting critical temperature ($T_c$) at ambient pressure, posing challenges in exploring the superconducting state. For instance, the upper critical field ($H_{c2}$) is small and thus difficult to measure accurately against other control parameters. Hence, enhancing superconductivity would facilitate $H_{c2}$ measurements, providing insights into key superconducting properties such as the dimensionality. In this letter, we report the tuning of the $T_c$ in RbV$_3$Sb$_5$ through the application of biaxial strain. Utilizing a negative thermal expansion material ZrW$_2$O$_8$ as a substrate, we achieve a substantial biaxial strain of $ε=1.50\%$, resulting in a remarkable 75\% enhancement in $T_c$. We investigate the $H_{c2}$ as a function of temperature, revealing a transition from multi-band to single-band superconductivity with increasing tensile strain. Additionally, we study the $H_{c2}$ as a function of field angle, revealing a plausible correlation between the $T_c$ enhancement and the change in dimensionality of the superconductivity under tensile strain. Further analysis quantitatively illustrates a transition towards two-dimensional superconductivity in RbV$_3$Sb$_5$ when subjected to tensile strain. Our work demonstrates that the application of biaxial strain allows for the tuning of both the $T_c$ and superconducting dimensionality in RbV$_3$Sb$_5$.

cond-mat.supr-con↗
arXiv2025

Bayesian shape-constrained regression for quantifying Alzheimer's disease biomarker progression

Several biomarkers are hypothesized to indicate early stages of Alzheimer's disease, well before the cognitive symptoms manifest. Their precise relations to the disease progression, however, is poorly understood. This lack of understanding limits our ability to diagnose the disease and intervene effectively at early stages. To provide better understanding of the relation between the disease and biomarker progressions, we propose a novel modeling approach to quantify the biomarkers' trajectories as functions of age. Building on monotone regression splines, we introduce two additional shape constraints to incorporate structures informed by the current medical literature. First, we impose the regression curves to satisfy a vanishing derivative condition, reflecting the observation that changes in biomarkers generally plateau at early and late stages of the disease. Second, we enforce the regression curves to have a unique inflection point, which enhances interpretability of the estimated disease progression and facilitates assessment of temporal ordering among the biomarkers. We fit our shape-constrained regression model under Bayesian framework to take advantage of its ability to account for the heterogeneity in disease progression among individuals. When applied to the BIOCARD data, the model is able to capture asymmetry in the biomarkers' progressions while maintaining interpretability, yielding estimates of the curves with temporal ordering consistent with the existing scientific hypotheses.

stat.AP↗
arXiv2025

ProtoBERT-LoRA: Parameter-Efficient Prototypical Finetuning for Immunotherapy Study Identification

Identifying immune checkpoint inhibitor (ICI) studies in genomic repositories like Gene Expression Omnibus (GEO) is vital for cancer research yet remains challenging due to semantic ambiguity, extreme class imbalance, and limited labeled data in low-resource settings. We present ProtoBERT-LoRA, a hybrid framework that combines PubMedBERT with prototypical networks and Low-Rank Adaptation (LoRA) for efficient fine-tuning. The model enforces class-separable embeddings via episodic prototype training while preserving biomedical domain knowledge. Our dataset was divided as: Training (20 positive, 20 negative), Prototype Set (10 positive, 10 negative), Validation (20 positive, 200 negative), and Test (71 positive, 765 negative). Evaluated on test dataset, ProtoBERT-LoRA achieved F1-score of 0.624 (precision: 0.481, recall: 0.887), outperforming the rule-based system, machine learning baselines and finetuned PubMedBERT. Application to 44,287 unlabeled studies reduced manual review efforts by 82%. Ablation studies confirmed that combining prototypes with LoRA improved performance by 29% over stand-alone LoRA.

cs.CL↗
arXiv2025

Mamba-VA: A Mamba-based Approach for Continuous Emotion Recognition in Valence-Arousal Space

Continuous Emotion Recognition (CER) plays a crucial role in intelligent human-computer interaction, mental health monitoring, and autonomous driving. Emotion modeling based on the Valence-Arousal (VA) space enables a more nuanced representation of emotional states. However, existing methods still face challenges in handling long-term dependencies and capturing complex temporal dynamics. To address these issues, this paper proposes a novel emotion recognition model, Mamba-VA, which leverages the Mamba architecture to efficiently model sequential emotional variations in video frames. First, the model employs a Masked Autoencoder (MAE) to extract deep visual features from video frames, enhancing the robustness of temporal information. Then, a Temporal Convolutional Network (TCN) is utilized for temporal modeling to capture local temporal dependencies. Subsequently, Mamba is applied for long-sequence modeling, enabling the learning of global emotional trends. Finally, a fully connected (FC) layer performs regression analysis to predict continuous valence and arousal values. Experimental results on the Valence-Arousal (VA) Estimation task of the 8th competition on Affective Behavior Analysis in-the-wild (ABAW) demonstrate that the proposed model achieves valence and arousal scores of 0.5362 (0.5036) and 0.4310 (0.4119) on the validation (test) set, respectively, outperforming the baseline. The source code is available on GitHub:https://github.com/FreedomPuppy77/Charon.

cs.CV↗
arXiv2025

Cross-Attention Fusion of MRI and Jacobian Maps for Alzheimer's Disease Diagnosis

Early diagnosis of Alzheimer's disease (AD) is critical for intervention before irreversible neurodegeneration occurs. Structural MRI (sMRI) is widely used for AD diagnosis, but conventional deep learning approaches primarily rely on intensity-based features, which require large datasets to capture subtle structural changes. Jacobian determinant maps (JSM) provide complementary information by encoding localized brain deformations, yet existing multimodal fusion strategies fail to fully integrate these features with sMRI. We propose a cross-attention fusion framework to model the intrinsic relationship between sMRI intensity and JSM-derived deformations for AD classification. Using the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, we compare cross-attention, pairwise self-attention, and bottleneck attention with four pre-trained 3D image encoders. Cross-attention fusion achieves superior performance, with mean ROC-AUC scores of 0.903 (+/-0.033) for AD vs. cognitively normal (CN) and 0.692 (+/-0.061) for mild cognitive impairment (MCI) vs. CN. Despite its strong performance, our model remains highly efficient, with only 1.56 million parameters--over 40 times fewer than ResNet-34 (63M) and Swin UNETR (61.98M). These findings demonstrate the potential of cross-attention fusion for improving AD diagnosis while maintaining computational efficiency.

eess.IV↗
arXiv2025

Discovery of a new phase in thin flakes of KV$_{3}$Sb$_{5}$ under pressure

We report results of magnetotransport measurements on KV$_3$Sb$_5$ thin flakes under pressure. Our zero-field electrical resistance reveals an additional anomaly emerging under pressure ($p$), marking a previously unidentified phase boundary $T^{\rm \ast}$($p$). Together with the established $T_{\rm CDW}(p)$ and $T_c(p)$, denoting the charge-density-wave transition and a superconducting transition, respectively, the temperature-pressure phase diagram of KV$_3$Sb$_5$ features a rich interplay among multiple phases. The Hall coefficient evolves reasonably smoothly when crossing the $T^{\rm \ast}$ phase boundary compared with the variation when crossing $T_{\rm CDW}$, indicating the preservation of the pristine electronic structure. The mobility spectrum analysis provides further insights into distinguishing different phases. Finally, our high-pressure quantum oscillation studies up to 31 T combined with density functional theory calculations further demonstrate that the new phase does not reconstruct the Fermi surface, confirming that the translational symmetry of the pristine metallic state is preserved.

cond-mat.supr-con↗
arXiv2024

Quantum Phase Transition as a Promising Route to Enhance the Critical Current in Kagome Superconductor CsV$_{3}$Sb$_{5}$

Developing strategies to systematically increase the critical current, the threshold current below which the superconductivity exists, is an important goal of materials science. Here, the concept of quantum phase transition is employed to enhance the critical current of a kagome superconductor CsV$_3$Sb$_5$, which exhibits a charge density wave (CDW) and superconductivity that are both affected by hydrostatic pressure. As the CDW phase is rapidly suppressed under pressure, a large enhancement in the self-field critical current ($I_{\rm c,sf}$) is recorded. The observation of a peak-like enhancement of $I_{\rm c,sf}$ at the zero-temperature limit ($I_{\rm c,sf}(0)$) centred at $p^*\approx 20$~kbar, the same pressure where the CDW phase transition vanishes, further provides strong evidence of a zero-temperature quantum anomaly in this class of pressure-tuned superconductor. Such a peak in $I_{\rm c,sf}(0)$ resembles the findings in other well-established quantum-critical superconductors, hinting at the presence of enhanced quantum fluctuations associated with the CDW phase in CsV$_3$Sb$_5$.

cond-mat.supr-con↗