arXiv subjects

Wei Lu

Publications and source records attributed to Wei Lu.

At least 19 recordsLinked to original sources

arXiv2026

Learning qBIC Resonances across Metasurface Families in Dielectric Fourier Space

Bound states in the continuum (BIC) metasurfaces are typically described by geometry-specific parameters, hindering cross-geometry comparison, while ultranarrow quasi-BIC (qBIC) features are easily diluted in full-spectrum learning. Here, we map 2,015 samples from seven dielectric metasurface families onto a shared reciprocal-lattice grid. Two fixed low-order Fourier descriptors capture within-branch resonance shifts, with mean $R^2$ values of 0.871-0.999. Field-level analysis of two representative branches shows that these shifts are consistent with the Maxwell-Fourier perturbation picture. A five-channel Fourier-space backbone models the broadband spectrum. A local resonance expert uses complex Fourier coefficients to parameterize the qBIC resonance through a differentiable Fano layer. The expert reduces resonance-position mean absolute error (MAE) from 3.2 to 0.95 nm and reduces the minimum-transmission MAE by a factor of 14 on a geometry-blocked test set. The same coordinate supports spectrum-to-structure reconstruction.

physics.optics↗

arXiv2026

Self-Evolution Reasoner: Continuous Optimization via Policy-Intrinsic Exploration and Exploitation

Reinforcement learning with verifiable rewards (RLVR) enables large language models (LLMs) to continuously self-improve their reasoning capabilities. Compared to RLVR's scalar rewards, recent self-supervised on-policy distillation methods leverage denser supervision signals. However, both paradigms suffer from exploration bottlenecks, struggling to escape local optima when the policy fails to sample valid solutions. Relying on external demonstrations can bypass this issue; however, the off-policy nature hinders genuine self-improvement and the mismatched-distribution reasoning can not be fully exploited. In this work, we propose Self-Evolution Policy Optimization (SEPO), a framework that drives continuous reasoning capability evolution via guided exploration and in-distribution exploitation. Specifically, SEPO utilizes labels to elicit correct reasoning from the policy, unlocking the exploration potential. By treating current policy conditioned on this in-distribution reasoning as a self-teacher, SEPO distills the newly discovered solution back into the policy. Through this iterative cycle of policy-intrinsic exploration and exploitation, SEPO achieves robust and continuous capability evolution. Across scientific, tool-use, and complex logical reasoning domains, SEPO consistently achieves the best performance under identical training time or steps, demonstrating superior efficiency and convergence ceilings. Notably, SEPO maintains stable improvements even in scenarios where baselines severely underperform or fail entirely, such as on complex tasks or with small-scale models. Our code is available at https://github.com/Eric8932/SEPO.

cs.LG↗

arXiv2026

Magnonic time crystal with lasing coherence

Discrete time crystals (DTC) can respond at an integer multiple of the drive period. Internal order-parameter measurements have advanced studies of this novel non-equilibrium phase of matter; however, turning this order into highly coherent, externally accessible radiation remains challenging: pump-driven nonlinear processes must supply sufficient gain to compensate dissipation while preserving subharmonic stability. Here we realize a room-temperature magnonic time crystal (MTC) with radiation of lasing coherence using a pump-induced magnon mode in yttrium iron garnet. A microwave beat drive produces robust period doubling over a finite parameter range, with spectral signatures of higher-order MTC at stronger driving. The emitted field exhibits linewidth narrowing to 260.0 Hz and a first-order correlation decay time of $995 μs$. Our model links this response to three-magnon confluence, yields a stability region consistent with experiment and predicts a possible transition to chaos. We further map the complex temporal dynamics into frequency-resolved nonlinear features for reservoir computing. These results bridge time-crystalline order and nonlinear information processing through intrinsic magnon-magnon interactions and lasing coherence.

cond-mat.mes-hall↗

arXiv2026

Why Do Agent Teams Fail Across Boundaries? A Failure Taxonomy for Agent-to-Agent Systems with Private States

Agents are increasingly interacting with other agents that act for different principals (people or organizations on whose behalf they work), such as a buyer's personal agent negotiating with an enterprise's seller agent. In such agent-to-agent (A2A) scenarios, each agent has its own environment and holds information and authority that the other cannot see. Moreover, the interests of each agent may be aligned (cooperative), partly aligned (mixed-motive), or opposed (competitive). Existing multi-agent system failure studies assume one orchestrator, shared workspace, or single principal. We ask how A2A systems fail when these assumptions are removed. We study more than 2,500 recorded runs of A2A systems built from different agent harnesses and models, spanning 488 scenarios across eight families (legal, CRM, security, commerce, etc.). From these traces, we derive $A^2$-Failures, a taxonomy of 20 A2A failure modes in four categories: (i) discovery and delegation, (ii) grounding and record consistency, (iii) verification, and (iv) conduct and fidelity. Three findings hold across every harness, model, and scenario studied: (1) discovery and interaction are driven by the phrasing of the task rather than the availability of peer agents; (2) no agent verifies a peer's factual claim on its own initiative, so an A2A system inherits the reliability of its least reliable participant unless the protocol or mechanism performs the check; and (3) the aggregate outcome score credits the agent completing the task while staying blind to the coordination, verification, and conduct failures beneath it, so A2A systems must be evaluated per agent, against each agent's own principal. We release the run logs behind the taxonomy at https://huggingface.co/datasets/a2a-salesforce/a2-failures.

cs.MA↗

arXiv2026

To Reason but Not to Stop: Understanding and Recovering Pruned Large Reasoning Models

Weight pruning can degrade the accuracy of large reasoning models (LRMs), yet it remains unclear whether it damages reasoning itself. We find that much of this degradation stems from a failure to stop rather than to reason. As sparsity increases, pruned LRMs increasingly fail to terminate within the token budget, often after the correct answer has appeared in their reasoning, instead falling into repetitive loops, even though they still recognize where the dense model would stop. We further find that the hidden-state difference between terminating and looping continuations is consistent across pruned checkpoints, suggesting a shared direction associated with committing to an answer. Building on this, we propose Termination-Guided Activation Steering (TGAS), a training-free method that adds this direction to the residual stream at inference time. Across six LRMs, three pruning pipelines, and math and code benchmarks, TGAS improves termination and overall accuracy while reducing repetitive reasoning and generation length, showing that this failure is actionable.

cs.CL↗

arXiv2026

DeltaTTG: Training Multi-Turn Agents with a Token-to-Go Progress Meter

Reinforcement learning with verifiable rewards (RLVR) for multi-turn agents relies on sparse terminal rewards that say little about individual steps. Within-group step comparisons depend on reliable state matching and sufficiently diverse sampled outcomes, while direct step scoring may require costly external evaluators, human annotations, or task-specific heuristics. We introduce DeltaTTG to learn a reusable progress signal through a shared progress meter, using low-cost supervision from diverse policies' recorded trajectories without task-specific progress rules. Its supervision target, Token-to-Go (TTG), the number of tokens remaining to be generated until task completion, serves as an observable proxy for remaining completion effort. During actor training, changes in the frozen meter's estimates redistribute outcome-derived advantages across steps without altering the reward, preserving each trajectory's advantage sign and token-average scale. On ALFWorld, PaperSearchQA, and BFCLv3, DeltaTTG improves the average score over GRPO by 5.7 and 7.2 points with Qwen3-4B and Llama-3.1-8B-Instruct, respectively, outperforms all compared RL baselines on average, and keeps actor-training GPU-hours close to GRPO. Further experiments show improved sample efficiency and out-of-domain performance, and support the meter's ability to distinguish step quality and guide test-time action selection across actor sizes. The code is available at https://github.com/Xuyan923r/DeltaTTG.

cs.LG↗

arXiv2026

Hierarchical Multi-Agent Reasoning Systems for Knowledge-Grounded Material Substitution

AI-assisted scientific discovery increasingly seeks to move beyond generating plausible ideas toward reproducible, constraint-aware workflows that can support practical engineering decisions. We present MARS, a hierarchical Multi-Agent Reasoning System for knowledge-grounded material substitution under scientific and manufacturing constraints. We apply MARS to replace per- and polyfluoroalkyl substances (PFAS), a demanding materials-design problem in which substitutes must satisfy functional performance, environmental, experimental, and processing requirements. MARS decomposes substitution into a closed-loop workflow comprising requirement analysis, grounded candidate generation, and manufacturability assessment with process design. The system integrates retrieval-augmented generation, knowledge-graph reasoning, a database of lab-available materials, persistent memory of intermediate results, and rejection-aware iterative feedback. In a PFAS-substitution case study, MARS generates structured material requirements, identifies an evidence-grounded substitute candidate, and develops a lab-scale manufacturing route. In a multi-seed ablation evaluation, MARS achieves the highest overall performance across all three subsystems, with gains that persist when the backbone model is held fixed. These results suggest that hierarchical decomposition, knowledge grounding, and manufacturability-aware refinement improve material-substitution reasoning beyond model capability alone.

cond-mat.mtrl-sci↗

arXiv2026

Evolution of Transversely Asymmetric Electron Beams in Hollow Plasma Channels

Plasma wakefield acceleration in hollow plasma channels has emerged as a promising approach for positron acceleration, since an electron beam can drive wakes with a transversely uniform accelerating field and no intrinsic defocusing force for positrons. Recently, it was proposed that a transversely asymmetric electron beam can excite quadrupole-dominated wakefield in a hollow channel, enabling the formation of accelerating and focusing fields suitable for positrons. However, the self-consistent evolution and stability of such asymmetric drivers, which are crucial for sustaining a usable wake over long distances, remain insufficiently understood. In this work, we investigate the evolution modes of wakefield driven by asymmetric electron beams in hollow plasma channels using fully three-dimensional particle-in-cell simulations. We identify two distinct unstable scenarios: a reversal of quadrupole field polarity and continuous penetration of the driver into the plasma wall. By analyzing the transverse dynamics of the driver and the restoring forces provided by the channel ions, we establish simple physical criteria that ensure stable propagation. These results clarify the fundamental constraints governing asymmetric-driver evolution, provide practical guidance for realizing long-lived, quasisteady wakes in hollow plasma channels, and indicate that the scheme can preserve mm$\cdot$mrad-level normalized positron emittance.

physics.plasm-ph↗

arXiv2026

Spatiotemporal imaging of microwave magnetic fields via magnonic coherent splitting

Spatiotemporal microwave magnetic-field imaging reveals current flow in high-frequency circuits and nonequilibrium spin dynamics, yet probes rarely combine calibrated spectral readout, optics-free operation and transient mapping at room temperature. Here ferrimagnetic order in yttrium iron garnet supports coherent coupling from a pump-induced magnon mode, converting target-field amplitude into a spectral splitting with all-microwave readout. Sampling the calibrated splitting over position and delay reconstructs spatiotemporal imaging of magnetic fields. Continuous-wave measurement reaches a sensitivity of 58 pT/$\sqrt{\mathrm{Hz}}$ and recovers phases across various powers. Combined with time-resolved frequency-comb spectroscopy, the method reconstructs transient fields with a 130-ns response time. Coplanar-waveguide imaging validates the magnetic selectivity of the mode-splitting readout, where measured maps agree with simulated magnetic-field distribution. In a microwave amplifier, our reconstruction resolves nonuniform switching dynamics and detects downstream field suppression from an open-contact fault. Magnonic coherent splitting provides a scalable route for optics-free imaging of spatiotemporal field evolution in functional devices.

cond-mat.mes-hall↗

arXiv2026

Breaking the Illusion of Review Reliability under Static Evaluation: SCOPE Fuzzing for LLM-based Scientific Reviewers

The rapid growth of submissions and reviewing workload has accelerated the use of large language models (LLMs) in peer review. Prior studies suggest that LLM-based reviewers can penalize content perturbations, such as overclaiming, indicating a certain degree of reliability. Yet these conclusions are largely based on a narrow set of perturbation strategies instantiated with static templates, providing limited evidence of actual reliability. In this paper, we construct a three-level evaluation framework covering perturbations to surface presentation, argumentative logic, and value judgment. Experiments on representative LLM-based reviewers reveal two limitations of static evaluation: stratified vulnerability, where perturbation effects depend on whether the paper's original review score is high or low, and perturbation undercoverage, where a single template misses vulnerabilities exposed by diverse realizations. To address these limitations, we propose SCOPE-Fuzzer, a strategy-aware fuzzer that combines feedback-driven strategy selection with adaptive mutation of paper content. By iteratively probing reviewers with dynamic perturbations, SCOPE-Fuzzer consistently uncovers vulnerabilities overlooked by static evaluation and other baselines.

cs.AI↗

arXiv2026

Multilingual Fine-Tuning via Localized Gradient Conflict Resolution

Cross-lingual versatility has become a defining capability of Large Language Models (LLMs). However, fine-tuning these models often induces negative interference, where gains in one language come at another's expense. To address this, we cast multilingual fine-tuning as a multi-objective optimization (MOO) problem. Specifically, we introduce Bucket-Level MOO, a scalable distributed framework that applies gradient-based MOO algorithms locally, within the parameter buckets of distributed training. This enables conflict-aware updates without ever materializing a full gradient vector for any objective, and improves multilingual performance over standard fine-tuning by up to $4.3$ points. Theoretically, we define bucketwise Pareto stationarity, show it is necessary for Pareto optimality under any fixed partition, and prove descent and convergence for representative Bucket-Level variants. Our work brings conventional MOO within reach for multilingual LLMs at a far lower cost. More broadly, Bucket-Level MOO is agnostic to both the objectives and the update rule, readily extending to other multi-task learning at scale.

cs.AI↗

arXiv2026

Safety Paradox: How Enhanced Safety Awareness Leaves LLMs Vulnerable to Posterior Attack

Large language models (LLMs) are rigorously aligned to refuse harmful requests, a process that inherently cultivates a latent capacity to evaluate and recognize unsafe content. In this work, we reveal that this advanced safety awareness inadvertently introduces a fatal vulnerability. We introduce Posterior Attack, a single-query jailbreak that bypasses guardrails by prompting the model to generate the exact harmful response its internal classifier would normally flag as unsafe. Through extensive empirical evaluation across 30 open-source LLMs (up to 35B parameters in size) and frontier models (e.g., GPT-5, Claude 4.6), we observe a striking phenomenon: models with superior safety-judgment capabilities are disproportionately more susceptible to this exploitation. To explain this, we formalize the Safety Paradox, analytically showing that monotonic improvements in safety alignment naturally amplify posterior vulnerability. Finally, we establish a causal link via reinforcement learning interventions, exemplifying that artificially degrading a model's safety judgment immunizes it against the attack, whereas enhancing judgment exacerbates the vulnerability. Our findings highlight potential flaws in current alignment paradigms, indicating that defense mechanisms may require further structural refinement.

cs.AI↗

arXiv2026

MARE: Multimodal Alignment and Reinforcement for Explainable Deepfake Detection via Vision-Language Models

Deepfake detection is a widely researched topic that is crucial for combating the spread of malicious content, with existing methods mainly modeling the problem as classification or spatial localization. The rapid advancements in generative models impose new demands on Deepfake detection. In this paper, we propose multimodal alignment and reinforcement for explainable Deepfake detection via vision-language models, termed MARE, which aims to enhance the accuracy and reliability of Vision-Language Models (VLMs) in Deepfake detection and reasoning. Specifically, MARE designs comprehensive reward functions, incorporating reinforcement learning from human feedback (RLHF), to incentivize the generation of text-spatially aligned reasoning content that adheres to human preferences. Besides, MARE introduces a forgery disentanglement module to capture intrinsic forgery traces from high-level facial semantics, thereby improving its authenticity detection capability. We conduct thorough evaluations on the reasoning content generated by MARE. Both quantitative and qualitative experimental results demonstrate that MARE achieves state-of-the-art performance in terms of accuracy and reliability.

cs.CV↗

arXiv2026

VikingRAG: Accurate and Token-efficient Retrieval-augmented Generation over Structured Documents

State-of-the-art retrieval-augmented generation (RAG) methods exploit document structures to acquire sufficient evidence, but often incur substantial token costs. To reduce structural-context tokens without compromising high RAG accuracy, we present {\sf VikingRAG}, a directory-aware semantic data management system that tightly integrates semantic and structural access to support structural-context-efficient, evidence-gap-driven multi-round retrieval. To further reduce token overhead of multi-round interaction, we materialize agentic multi-round retrieval traces as experience edges, and reuse these edges for similar queries, avoiding repeated multi-round exploration. To additionally reduce token costs when agentic multi-round retrieval is unnecessary, we introduce an adaptive escalation strategy that answers from one-round experience-augmented retrieval when the evidence is sufficient, and invokes agentic multi-round retrieval only otherwise. Experiments on real datasets show that the base system {\sf VikingRAG} matches high accuracy of state-of-the-art methods while consuming only 11.6\%--51.9\% of their tokens. With retrieval-trace reuse and adaptive escalation, token costs drop to 5.1\%--32.5\% while maintaining competitive accuracy and practical document-storage performance, showing the utility of this work for emerging AI knowledge bases.

cs.IR↗

arXiv2026

DBRepro: Automated Database Synthesis via a Hybrid Constraint-Solving Approach for Reproducing Slow Queries

Slow queries frequently cause severe performance bottlenecks in database management systems. Diagnosing their root causes online risks exacerbating resource contention, while data privacy regulations often prohibit copying production data to test environments. Synthesizing a proxy database from non-intrusive metadata that induces the query optimizer to generate the same physical execution plans is therefore critical for offline diagnosis. High-fidelity reproduction requires preserving global statistical distributions while enforcing exact local cardinalities. Existing data-driven and workload-aware approaches cannot satisfy both requirements simultaneously. We present DBRepro, an automated end-to-end framework that formulates database generation as a constrained distribution synthesis problem. DBRepro initializes a global distribution from lightweight column statistics, extracts execution constraints from target queries, and progressively adjusts the distribution to satisfy these constraints while preserving the global distribution. Experiments on TPC-H and SSB show that DBRepro reduces cardinality error by up to 20.3% over a data-driven baseline while maintaining identical plan consistency. Compared with a workload-aware baseline, it reproduces 15% more consistent execution plans and reduces latency proportion error by 21.5%. We further validate DBRepro on a nearly 1 TB real-world dataset managed by KingbaseES, where it reproduces the execution performance of complex slow queries with high fidelity.

cs.DB↗

arXiv2026

DBcover: A White-box SQL Test Generation Framework for Coverage Improvement

Relational Database Management Systems (RDBMSs) are the backbone of modern data-intensive applications, making reliability and robustness critical. However, achieving high coverage in RDBMS testing remains challenging because of large codebases and complex execution logic. Traditional fuzzing relies on random SQL generation and cannot capture the correspondence between SQL inputs and internal execution paths, while symbolic execution suffers from prohibitive cost and scalability limitations. We propose DBcover, an LLM-driven white-box SQL test generation framework based on contextual reasoning. DBcover uses lightweight dynamic analysis to extract SQL-to-path correspondence and call graphs as global context, and collects source-level information around target functions as local context. These contexts are organized in a unified knowledge graph for efficient retrieval and reuse. DBcover then performs two-phase test generation: it first selects a semantically relevant seed whose execution path is close to the uncovered target, and then guides the LLM with global and local context to generate SQL test cases that trigger previously uncovered code regions. Experiments show that DBcover achieves 80.1% and 82.3% coverage on PostgreSQL and MySQL, and is also effective on the enterprise RDBMS KingbaseES, demonstrating its practical applicability to closed-source systems.

cs.DB↗

arXiv2026

Field deployment of a laser wakefield accelerator for on-site application

Successive innovations in particle accelerators have continually expanded the frontiers of scientific discovery. Laser wakefield accelerators promise to transform science, medicine, and industry, yet moving them from laboratory demonstrations to reliable real-world operation has remained a central, long-standing challenge. Here we report a field-deployable system that produced 100-MeV-class electron beams with 1%-level energy stability during 72 hours of continuous operation and supported routine full-power use throughout a seven-month field trial in an industrial setting. Applied to in situ micro-nondestructive testing, the system generated tens-of-MeV bremsstrahlung X-rays that enabled three-dimensional microtomography of dense materials at sub-50-μm spatial resolution and revealed 100-μm-scale internal defects in large composite structures, extending the capabilities beyond those of existing high-energy X-ray sources. These results mark a transition of laser wakefield acceleration from laboratory proof of concept toward practical deployment in scientific and industrial applications.

physics.acc-ph↗

arXiv2026

Bright dual-pulse betatron X-ray generation from a laser wakefield accelerator

Pump-probe experiments using dual ultrashort X-ray pulses provide unique opportunities for resolving non-equilibrium dynamics initiated by intense X-ray excitation. Betatron radiation from laser wakefield accelerators offers femtosecond duration, micrometer-scale source size, and intrinsic synchronization with the driving laser, making it a promising candidate for compact ultrafast X-ray sources. Here, we experimentally demonstrate a high-flux, dual-pulse betatron X-ray source based on a density-tailored gas-mixture target. Two electron bunches are generated within a single plasma wakefield through ionization-induced and shock-front-triggered injection, subsequently producing twin X-ray pulses. The measured electron spectra and dual-component X-ray angular profiles, together with particle-in-cell simulations, identify the contributions of the two electron populations to the radiation. The total X-ray photon yield reaches the level of 10^{10} photons per shot with a 40-TW laser system. These results establish a compact, single-stage route toward high-flux dual-pulse betatron sources for laboratory-scale ultrafast X-ray spectroscopy.

physics.plasm-ph↗