arXiv subjects

Lei Wang

Publications and source records attributed to Lei Wang.

At least 145 records · Page 8Linked to original sources

arXiv2026

Inference of Component Effect on System Performance

In a computer system, multiple components--such as the CPU, memory, and others--work together as a system whose performance can be directly measured. However, the effect of a component under investigation (CUI), e.g., CPU, on system performance cannot be directly measured and can only be inferred. Accurately inferring CUI effect on system performance is a critical issue. Our experiments reveal that the general-purpose rigorous methodologies, like Design of Experiments (DoE), Randomized Controlled Trials (RCTs), and a single-purpose empirical methodology, like SPEC CPU2017, can not address this issue effectively and efficiently. We propose a rigorous methodology to address this issue: First, we identify a self-contained system (SCS) under the context of which we can completely understand how CUI and other essential components affect the system performance, and then we use a structural causal model methodology to represent and infer the causal effect of CUI on the system performance. We utilize this methodology and verify its correctness in the context of CPU design and evaluation. Through theoretical analysis and pioneering controlled experiments, we systematically compare our methodology against three established methodologies: SPEC CPU2017, DoE, and RCTs. The results show that our methodology can achieve its goal effectively and efficiently, whereas others exhibit inherent limitations.

cs.PF↗

arXiv2026

Irreversibility Enhances Quantum-Enhanced Markov-Chain Monte Carlo

Detailed balance underlies conventional Markov-chain Monte Carlo (MCMC) algorithms. Yet in classical systems, breaking detailed balance generates irreversible probability currents and can accelerate sampling. Whether irreversibility can similarly enhance quantum MCMC remains an intriguing question. Here we show that irreversibility provides a new route to improving the recent quantum-enhanced MCMC (QEMC), which combines quantum proposals with classical acceptance. By introducing state-dependent proposals that break detailed balance while preserving the target stationary distribution, we develop an irreversible quantum-enhanced Monte Carlo (IQEMC). Guided by Landau-Zener transitions, IQEMC promotes large energy descents from high-energy states while maintaining stable transitions near low-energy states. On spin-glass benchmarks, IQEMC outperforms QEMC without increasing computational complexity and, unlike the annealing baseline, exhibits a spectral gap that increases with system size and annealing speed. These results establish irreversibility as a physically grounded mechanism for enhancing quantum MCMC.

cond-mat.stat-mech↗

arXiv2026

Anomalous charge density wave in a two-dimensional superatomic superconductor

The spatial modulation of electron density into a wave-like pattern, known as charge density wave (CDW), represents a fundamental quantum state that often coexists with superconductivity, quantum Hall states, axion insulating phases and etc. Conventional CDWs are mediated by longitudinal acoustic phonons, exhibit picometer-scale lattice distortions ($10^{-12}$--$10^{-11}$ m), and typically vanish approaching the atomic limit. Here, we report a series of anomalous CDW behaviors in the 2D superatomic superconductor Au$_6$Te$_{12}$Se$_8$. Remarkably, its CDW is governed by transverse phonons, accompanied by an extraordinarily high real-space displacement of $\sim 4$ Ångström. Furthermore, we observe an exotic dimensional response persisting up to micrometer-scale thickness, a regime where other materials are already considered as bulk. Through liquid helium-temperature transmission electron microscopy, ultrafast pump-probe spectroscopy and transport measurements, we demonstrate a dramatic enhancement of the CDW transition temperature ($T_{\text{CDW}}$) from $<2$ K in the bulk to 110 K in approaching the ``superatomic limit''. Our findings not only reveal novel facets of both CDW and superatomic materials, but the competition between this anomalous CDW and superconductivity opens avenues for exploring unconventional electron-phonon interactions.

cond-mat.supr-con↗

arXiv2026

LaST-HD: Learning Latent Physical Reasoning from Scalable Human Data for Robot Manipulation

Human-hand demonstrations provide a direct and scalable source of physical interaction data for robot learning. While manual retargeting is indispensable for establishing kinematic action correspondence across different morphologies, robust transfer requires going beyond geometry to address the underlying alignment of physical dynamics between human and robot manipulation. To address this, we introduce LaST-HD, a novel human-to-robot action learning paradigm that extends reasoning-before-acting VLA by aligning human-hand and robot demonstrations in a shared latent reasoning space. Rather than mimicking human kinematics, LaST-HD trains an auxiliary action-conditioned world model on unpaired human-hand and robot trajectories to synthesize unified latent targets. After aligning cross-embodiment representations in this shared forward-dynamics space, these targets supervise LaST-HD's latent reasoning process, enabling it to internalize shared physical dynamics and drive efficient human-hand action learning. Moreover, we develop Out-of-Lab (OOL) Glove, a low-cost motion-capture glove tailored to LaST-HD for human-hand data collection. The captured human data provide precise keypoints and serve as universal action supervision across grippers and dexterous hands. Armed with the aligned latent space and high-fidelity human-hand data, we develop a progressive mixed-to-human training recipe comprising mixed human-robot co-training and human-hand online correction post-training. Through mixed co-training, LaST-HD improves generalization to novel objects, scenes, and positions using only human-hand demonstrations. With online correction, LaST-HD further adapts to novel environments and achieves over 90\% accuracy using only 20 minutes of OOL glove data.

cs.RO↗

arXiv2026

Vibe Calibration: Autonomous Bring-up of a 112-Qubit Superconducting Quantum Processor by a Skill-Orchestrating Language Agent

Superconducting quantum computing is one of the most mature solid-state platforms for quantum computation, with processors exceeding one hundred qubits. Yet further scaling toward fault-tolerant quantum computing is increasingly constrained by calibration complexity. Conventional scripts are brittle to anomalous signals, and expert judgment is bounded by cognitive bandwidth and serial operation time, failing to keep pace with system scale. Here we report Vibe Calibration, an autonomous calibration system orchestrated by large language model agents, which distills expert tacit knowledge into reusable Skills. Each Skill is organized as a decision tree that packages parameterized measurement commands, quantitative acceptance criteria, and audit records, enabling autonomous execution and self-healing. We capture this knowledge through a three-phase human-in-the-loop distillation process and fine-tune a large language model on validated trajectories. On a 112-qubit processor with frequency-tunable transmons, the system autonomously completes calibration of 108 out of 112 qubits in 4.7 hours, achieving a 4--5$\times$ speedup over manual calibration of the full 112 qubits. A cross-validated comparison with expert manual calibration on a 16-qubit subset shows agreement on 14 out of 16 qubits. More importantly, the model demonstrates transferable calibration workflows across devices. While low-level control scripts require minor interface adaptation for different hardware platforms, the core decision logic and task orchestration generalize to new processors, demonstrating a reusable laboratory interface rather than a memorized script.This work demonstrates, for the first time, fully autonomous calibration of a hundred-qubit superconducting processor through reusable and auditable Skills, removing a critical barrier to scalable quantum hardware operation.

quant-ph↗

arXiv2026

Structured Hyperedge Adaptation for Parameter-Efficient Fine-Tuning of Vision Transformers

Parameter-efficient fine-tuning (PEFT) has become a practical solution for adapting large pretrained vision transformers (ViTs) to downstream tasks while updating only a small subset of parameters. However, existing adapter-based methods perform adaptation independently for each token, implicitly assuming that token refinements should be learned in isolation. This token-wise formulation overlooks the structured relationships among tokens that naturally arise in visual scenes, potentially leading to redundant updates and spatially inconsistent feature refinement. In this work, we revisit the design of parameter-efficient adapters and propose to perform adaptation in hyperedge space rather than token space. We introduce HyperAdapter, a hypergraph-based adapter architecture that enables structured, group-aware adaptation through soft token routing. HyperAdapter constructs a soft hypergraph over ViT tokens using prototype-based assignments, aggregates token features into latent hyperedge representations, applies lightweight bottleneck adaptation at the hyperedge level, and diffuses the resulting updates back to tokens via the hypergraph incidence structure. This design injects an explicit structural inductive bias into PEFT while preserving the modularity and efficiency of standard adapters. Extensive experiments across diverse visual benchmarks demonstrate that structured hyperedge adaptation consistently outperforms strong PEFT baselines under comparable parameter budgets, with particularly pronounced gains on tasks requiring structured reasoning. Our results suggest that the choice of adaptation space is a critical yet underexplored dimension in parameter-efficient transfer for ViTs.

cs.CV↗

arXiv2026

Trip+: Benchmarking Agents in Personalized Interactive Travel Planning

Interactive travel planning has become a popular use case for language models. Agents are deployed to manage evolving preferences and unexpected disruptions over multiple turns. Such settings require models to make complex, profile-conditioned planning decisions. However, existing benchmarks often evaluate feasibility, personalization, or interaction in relatively isolated settings. We therefore introduce Trip+ to measure the ability of agents to plan travel holistically. In Trip+, given traveler profiles and dynamic interactions, agents must generate and revise minute-level itineraries. End-to-end traveler experiences are evaluated via an LLM-based simulator, enabling the assessment of subjective metrics like fatigue. Our scenarios range from simple request resolutions to complex environment-driven replanning. We evaluate 18 LMs and find a consistent gap in experiential quality. Models favor technically feasible but exhausting itineraries that diverge sharply from profiled traveler preferences.

cs.AI↗

arXiv2026

Terahertz frequency upconversion by coherently driving charge dynamics in the InSb/CdTe heterostructure

We investigate terahertz (THz) harmonic generation in the InSb/CdTe heterostructure, demonstrating, for the first time, efficient in-plane magnetic field-induced second-harmonic generation (SHG). We also achieve significant third-harmonic generation (THG), rivalling Dirac materials such as graphene and Cd3As2. Our theoretical analysis identifies the primary SHG mechanism as the orbital-Zeeman correction to Drude conductivity, while the dominant THG contribution also shows Drude-like behavior. The results provide a general route to efficient THz harmonic generation in high mobility materials.

cond-mat.mtrl-sci↗

arXiv2026

EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic Environments

Large language model (LLM) agents have achieved strong performance on a wide range of benchmarks, yet most evaluations assume static environments. In contrast, real-world deployment is inherently dynamic, requiring agents to continually align their knowledge, skills, and behavior with changing environments and updated task conditions. To address this gap, we introduce EvoArena, a benchmark suite that models environment changes as sequences of progressive updates across terminal, software, and social domains. We further propose EvoMem, a patch-based memory paradigm that records memory evolution as structured update histories, enabling agents to reason about environmental evolution through changes in their memory. Experiments show that current agents struggle on EvoArena, achieving an average accuracy of 39.6% across evolving terminal, software, and social-preference domains. EvoMem consistently improves performance, yielding an average gain of 1.5% on EvoArena and also improving standard benchmarks such as GAIA and LoCoMo by 6.1% and 4.8%. Beyond individual tasks, EvoMem further improves chain-level accuracy by 3.7% on EvoArena, where success requires completing a consecutive sequence of related evolutionary subtasks. Mechanistic analysis shows that EvoMem improves evidence capture in the memory, indicating better preservation of complete evolving environment states. Our results highlight the importance of modeling evolution in both evaluation and memory for reliable agent deployment.

cs.CL↗

arXiv2026

ARC-Fi: Exploiting Antenna Spatial Diversity for Label-Efficient Domain Generalization in Wi-Fi Sensing

Wi-Fi sensing systems are severely hindered by domain shifts when deployed in unseen real-world environments. While existing methods attempt to tackle this through Unsupervised Domain Adaptation (UDA) or Domain Generalization (DG), they critically rely on either inaccessible target data or prohibitively expensive, massive labeled source datasets. In practice, collecting abundant unlabeled Channel State Information (CSI) is feasible, whereas manual labeling is severely constrained. This realistic dilemma necessitates Semi-Supervised Domain Generalization (SSDG). To this end, we propose ARC-Fi, the first dedicated SSDG framework for Wi-Fi sensing. Directly applying conventional contrastive learning to CSI data inevitably triggers paradigm-specific "shortcut learning," causing models to memorize environmental backgrounds rather than gesture dynamics. To overcome this, ARC-Fi introduces a physics-informed data augmentation strategy: the Antenna Response Consistency (ARC) module. ARC exploits the intrinsic spatial diversity of multi-antenna systems, treating signals from co-located antennas as naturally semantics-preserving augmented views to explicitly block environmental shortcuts. Furthermore, we introduce a unified Semi-Supervised Contrastive Objective that leverages scarce labels and reliable pseudo-labels to align cross-domain features, effectively preventing the blind repulsion of same-class instances. Extensive experiments on the Widar and CSIDA datasets demonstrate that ARC-Fi establishes a new state-of-the-art, significantly outperforming existing UDA, DG, and SSDG methods. Ultimately, this work provides a physics-grounded, label-efficient solution, advancing the scalable deployment of robust real-world Wi-Fi sensing systems. Code is available at: https://github.com/KaoruMiyazono/UniCrossFi.

cs.LG↗

arXiv2026

Collective relaxation eigenmodes and anisotropic magnon thermal transport in $α$-MnTe

The intrinsic magnon thermal transport in the altermagnet $α$-MnTe is studied by solving the three-dimensional linearized Boltzmann transport equation with the four-magnon collision matrix. Diagonalizing the collision matrix gives direct access to the relaxation eigenmodes beyond the relaxation-time approximation. We show that Umklapp scattering lifts the momentum-related zero modes of the Normal-only collision operator and substantially modifies the low-lying relaxation spectrum. Using the same collision matrix, we compute the magnon thermal conductivity tensor. The full linearized Boltzmann transport equation result exceeds the relaxation-time approximation by more than an order of magnitude at low temperature and reveals a strong transport anisotropy, with the out-of-plane thermal conductivity remaining larger than the in-plane component over the studied temperature range. A mode-resolved analysis shows that the dominant heat-carrying modes retain momentum-like character inherited from the Normal-only zero modes, and that the larger out-of-plane conductivity mainly originates from the stronger out-of-plane group-velocity contribution, rather than from a large difference in relaxation lifetimes.

cond-mat.mtrl-sci↗

arXiv2026

Does Capability Transfer to Subjective Behavior -- and Would Our Instruments Tell Us? A Self-Evolving, Trust-by-Construction Evaluation Paradigm

Benchmarking is mature where answers are verifiable -- math, code, reasoning -- but the fastest-growing uses of LLMs are subjective and human-facing: companionship, emotional support, counseling. There the default validity test, correlating a metric to human judgment, has no stable anchor: inter-rater agreement is low, structured by annotator identity, barely reproducible, and length-biased. So we cannot answer the question that matters: does capability that scales on objective benchmarks transfer to subjective behavior, and would our instruments even tell us if it did not? We build an instrument for this regime and report what it reveals at the frontier. We contribute, first, a self-evolving instrument that selects and then authors its own behavioral dimensions under a multiplicative anti-gaming fitness, self-halting when it stops improving; second, a trust-by-construction paradigm that earns belief through three certificates established without a human gold standard, where human raters saturate (rho ~ 0.45); and third, the finding it makes visible -- capability transfer is dissociable. Across 49 models, 8 families, and 24 months, subjective behaviors are where objective-benchmark scaling fails to carry over: the sharpest case, advice-restraint (knowing when not to give advice), is the frontier's universal-lowest dimension, and at gpt-4.1->gpt-5 it ran backwards while the aggregate score hid it -- a regression one instruction recovers. Warm restraint is moved by model generation, not by raw scale, MoE width, inference budget, or reasoning mode; the open-weight Pareto frontier matches closed flagships at ~10-80x lower per-call cost; and four judge families replicate the rubric on held-out human ESConv conversations. Data, code, the locked rubric, and judge prompts will be released upon publication.

cs.CL↗

arXiv2026

Video Understanding by Design: How Datasets Shape Video Models

Research in video understanding has advanced rapidly, driven by increasingly diverse datasets and more powerful model architectures. While existing surveys typically organize progress by tasks, benchmarks, or model families, they provide limited insight into why particular architectures emerged and succeeded. In this survey, we argue that the evolution of video understanding is fundamentally shaped by dataset structure. We present a dataset-centric perspective that connects dataset structure, inductive biases, and architectural design within a unified framework. We show that different datasets require models to capture specific invariances and capabilities, such as robustness to viewpoint changes, sensitivity to temporal ordering, reasoning over long-range dependencies, relational interactions, and cross-modal alignment. These requirements naturally give rise to inductive biases, i.e., architectural assumptions that favor particular patterns of reasoning and generalization. From this perspective, milestone architectures, including two-stream networks, 3D CNNs, temporal models, transformers, graph-based methods, and multimodal foundation models, can be understood as architectural responses to the challenges posed by evolving datasets. Building on this framework, we systematically analyze how dataset characteristics have shaped architectural innovation across video understanding tasks and discuss the representational biases induced by different data regimes. By unifying datasets, inductive biases, and architectures into a coherent perspective, this survey offers both a retrospective explanation of the field's evolution and a forward-looking roadmap toward general-purpose video understanding systems. Code and dynamic video visualizations of dataset-induced biases are available at https://time.griffith.edu.au/paper-sites/video-understanding/.

cs.CV↗

arXiv2026

Benchmarking LLMs for Community Governance Simulation with Life-history Narratives

Effective community governance hinges on understanding what specific residents think and need. Recent work has used large language models (LLMs) to simulate human respondents, offering a scalable, reproducible way to study human attitudes and behaviors at low cost. However, these studies typically prompt the model with just a few demographic variables (age, gender, income), simulating only general role types. This is insufficient for community governance, where decisions depend on the views of specific residents. We bridge this gap with an integrated research framework covering dataset, benchmark, algorithm, and system. The dataset comprises approximately 1.2 million characters of first-person narrative collected through two-hour semi-structured interviews with each of 92 residents in an urban community, organized around nine community-governance domains. The benchmark probes 18 mainstream LLMs across four prompting strategies and shows that adding rich life-history profiles meaningfully raises fidelity above the no-profile baseline, but this gain comes with more input tokens per call from the longer prompts they require. The algorithm, curriculum-LoRA, is a parameter-efficient personalization framework that, by closing this fidelity-cost gap, matches the strongest baseline's fidelity at roughly 10x lower per-call cost and Pareto-dominates every configuration tested. The system integrates curriculum-LoRA into a closed-loop policy-evaluation pipeline. Together, these results bring individual-level LLM-based resident simulation within reach of resource-constrained local administrations, enabling community-governance decisions to be systematically pre-evaluated in silico before real-world deployment.

cs.CY↗

arXiv2026

Neural Autoregressive Control Variates for the Quantum Monte Carlo Sign Problem

We train a pair of autoregressive models to construct zero-mean control variates to mitigate the sign problem in quantum Monte Carlo simulations. The two autoregressive networks are confined to the positive- and negative-sign sectors with strictly disjoint support, and each is exactly normalized over its sector. Their difference is therefore structurally zero-mean, providing an unbiased auxiliary observable whose correlation with the sign estimator controls the variance reduction. We implement the method within the stochastic series expansion framework, which we extend to frustrated lattices by developing an incremental loop-topology update. Sign-ergodic sampling is achieved through a twist channel, which is the unique sign-changing mechanism on non-bipartite lattices. We implement the control variates as autoregressive transformers with an end-of-sequence parity mask that enforces exact sign-sector resolution, while the incremental loop-count change and cumulative frustration parity are incorporated as topological features. On the triangular-lattice Heisenberg antiferromagnet, we benchmark the method in the small-$N$ limit. The control variate reduces the standard error of the average sign by up to an order of magnitude and that of the energy estimator by a factor of three to five, remaining effective even when the average sign drops below $10^{-3}$. This work lays out the framework and provides a proof-of-principle demonstration that autoregressive control variates can effectively mitigate the sign problem. Scaling to larger systems with physics-informed architectures is the subject of future work.

cond-mat.str-el↗

arXiv2026

MemoGen: Can Past Experience Improve Future Text-to-Image Generation?

Modern text-to-image models have achieved strong visual synthesis, yet remain unreliable when prompts require implicit visual constraints, relational reasoning, or external knowledge. Existing retrieval-augmented and agentic generation methods mitigate this issue by acquiring external knowledge, references, or refined prompts for the current request, yet they typically treat each generation as an isolated episode and do not systematically preserve past successes or failures for future use. In this work, we ask whether a text-to-image system can continually improve from its own generation experience without updating the underlying generator. We propose MemoGen, a training-free framework that augments existing image generators with an agentic evolution layer. For each task, MemoGen explicitly infers visual requirements, retrieves external evidence and references when necessary, translates them into executable generation constraints, evaluates the generated result, and stores task understanding, reference choices, visual feedback, successful strategies, and failure lessons as reusable experience memory. Across evolution rounds, the agent retrieves relevant experience to improve similar future generations, selectively repairing previously failed cases while preserving successful ones, thereby enabling test-time self-evolution without parameter updates. Extensive experiments on knowledge-intensive and reasoning-oriented benchmarks demonstrate the effectiveness of this paradigm: after only two evolution rounds, MemoGen built upon the open-source Qwen-Image backbone surpasses strong proprietary systems such as Nano Banana Pro and GPT-Image-1 on WISE and Mind-Bench, showing that explicit experience memory can serve as a powerful continual learning signal for reliable text-to-image generation.

cs.CV↗

arXiv2026

Tailoring pure valley-Zeeman spin-orbit coupling in WSe$_2$-encapsulated monolayer graphene

Engineering proximity effects in twisted van der Waals heterostructures offers a powerful platform for designing electronic properties. While theoretical predictions of quantum interference in transition metal dichalcogenide-encapsulated graphene can selectively control the spin-orbit coupling component, experimental realizations have remained elusive. Here, we report pure valley-Zeeman spin-orbit coupling in monolayer graphene, achieved by encapsulation between two parallel twisted WSe$_2$ monolayers. We observed a symmetry-enforced reordering of Landau levels, which is driven by the competition between the fixed valley-Zeeman energy and the magnetic-field-dependent cyclotron energy. This reordering is characterized by a transition from symmetry-broken states in the quantum Hall effect to a restored fourfold degeneracy with integer or half-integer quantum Hall sequences. We also demonstrate the ability to completely quench the proximity spin-orbit coupling by tuning the encapsulated geometry.

cond-mat.mes-hall↗

arXiv2026

Enhanced superconductivity in atomically thin noble metals: From quantum confinement to interface-induced Lifshitz transition

Unlocking superconductivity in intrinsically non-superconducting noble metals (Au, Ag, Cu) represents a fundamental challenge in low-dimensional physics. While quantum confinement in the atomically thin limit is known to trigger emergent superconductivity, strategies to amplify this marginal effect to experimentally accessible temperatures remain a key open question. Using first-principles calculations, we establish a unified framework linking intrinsic confinement effects with interface engineering in noble metal films. We reveal that intrinsic superconductivity is element-specific: it is suppressed in Ag by a stiff phonon spectrum, but emerges in trilayer Cu ($T_{\rm C} \approx 0.78$ K) and pentalayer Au ($0.63$ K) driven by confinement-induced density-of-states (DOS) enhancement and phonon softening, respectively. In h-BN/Cu(111) heterostructures, $T_{\rm C}$ is critically dictated by the interfacial stacking configuration. We identify the thermodynamically stable N-bonded interface as a reliable platform for accessible superconductivity ($T_{\rm C} \approx 3.23$ K), whereas manipulating the system into a metastable B-bonded configuration boosts $T_{\rm C}$ to $7.00$ K. This enhancement originates from a B-bonded-induced Lifshitz transition, where the Fermi surface forms a tangential contact with the Brillouin zone boundary at the M point, enhancing electron-phonon coupling beyond DOS effects. Our work unifies the understanding of intrinsic two-dimensional superconductivity with atomistic interface design, offering a blueprint for functionalizing noble metals as emergent superconductors.

cond-mat.supr-con↗