arXiv subjects

Dawei Wang

Publications and source records attributed to Dawei Wang.

At least 19 recordsLinked to original sources

arXiv2026

Experimental demonstration of phase-conjugated twin-wave coherent reception over 100 km with injection locking assisted phase stabilization

Recovering complex optical signals with high sensitivity is central to coherent communication and sensing. Beyond conventional homodyne and heterodyne reception, phase-conjugate twin-wave encoding and phase-sensitive joint detection offer an alternative strategy with enhanced sensitivity. Extending this approach to long-distance transmission of high-order-modulated signals is challenging because optical phase coherence must be maintained at the receiver. Here we experimentally demonstrate injection-locking-assisted phase-conjugated twin-wave reception of 200-MBaud signals with modulation formats up to 256-QAM over 100 km of optical fiber. Optical injection locking regenerates the local oscillator (LO) from a reference transmitted through the 16 dB-loss link, while pilot-assisted optical feedback stabilizes the relative signal-LO phase, enabling coherent combining of the phase-conjugated twin waves at a single balanced photodetector. At equal received signal-photon number per quadrature in each data-carrying optical mode, the measured PCTW signal-to-noise ratio (SNR) exceeds the ideal single-mode homodyne standard quantum limit (SQL) benchmark. The SNR gains over the measured homodyne and conventional heterodyne results are about 3 dB and 6 dB, respectively. Constellation and bit-error-rate measurements show that this gain improves high-order data recovery relative to heterodyne reception. An ideal Gaussian-noise model predicts a photon information efficiency of 1.54 bits per received signal photon for uniform 16-QAM. These results establish an injection-locking-assisted coherent receiver architecture that implements joint detection of phase-conjugated twin waves over 100 km of fiber with enhanced sensitivity, providing a practical route for photon-efficient coherent communication.

physics.optics↗

arXiv2026

ROAR: Benchmarking LMM Agents under Real-Time Constraints in Multiplayer Action-Shooter Games

Large multimodal models (LMMs) are increasingly deployed as agents that perceive, reason, and act in interactive multi-agent environments. Yet most existing benchmarks for LMM-based agents in multi-agent settings remain turn-based: the environment and other agents pause while an agent performs inference. Such protocols obscure a fundamental challenge of real-time constraints, where the world continues to evolve and delayed decisions may become outdated before execution. Motivated by this limitation, we introduce ROAR, a multiplayer action-shooter game environment for evaluating LMM-based agents in multi-agent settings under real-time constraints. ROAR covers cooperative, competitive, mixed cooperative-competitive, and basic single-agent scenarios. We further propose a Multi-Agent Asynchronous Gym interface that decouples environment evolution from agent reasoning by running the environment and each agent in separate threads, enabling the environment to advance at fixed time intervals without waiting for LMM inference and allowing agents to reason independently without blocking one another. ROAR also features convenient deployment, multimodal observations, and customizable levels. ROAR offers a practical testbed for advancing LMM-based agents beyond static, turn-based evaluation toward real-time multi-agent decision-making.

cs.MA↗

arXiv2026

Mining the Self-Rewriting System: A Vision for Data Mining on AI-Native Infrastructure

Classical data mining treats data as generated outside the learner. Emerging AI-native infrastructure invalidates this assumption: as agents observe system state, act, update retrieval stores and models, and promote new artifacts into the serving path, their own decisions reshape the future logs, traces, corpora, and evaluation records used for later training and release. We call this new research object a self-rewriting system and the resulting system-induced distribution shift endogenous drift. Our central, falsifiable claim is that cross-surface causes are not identifiable in general from marginal per-surface monitoring: different coupled failures can produce the same individual signals while differing in their joint behavior, requiring joint analysis for reliable attribution. This agenda comprises four foundational research problems: fusing heterogeneous drift signals into calibrated gating and attribution (C1); mining agent traces whose activity alphabet evolves (C2); maintaining knowledge fabrics through online reindexing, faithfulness, and compression decisions (C3); and safely promoting concurrent learning streams through one shared evaluation gate (C4). These remain fragmented across concept drift, performative prediction, process mining, streaming retrieval-augmented generation, and MLOps, without a unified data-mining agenda. We outline public benchmarks and evaluation criteria, using AI-native telecom only as a motivating exemplar with analogues in finance, software CI/CD, and clinical decision support.

cs.LG↗

arXiv2026

Delta-sigma noise shaping for low-resolution quantization in Gaussian-modulated continuous-variable quantum key distribution

Low-resolution digital-to-analog conversion can simplify continuous-variable quantum key distribution (CVQKD) transmitters, but coarse quantization increases excess noise and can generate data-correlated spectral replicas. We demonstrate digital delta--sigma modulation (DSM) for low-resolution waveform generation in Gaussian-modulated CVQKD. In a 500-MBaud CVQKD system over 20~km of fiber, 3-bit second-order DSM achieves an estimated asymptotic secret key rate of 46.38~Mbps at an oversampling ratio of 12, whereas uniform quantization yields no positive rate. With 4-bit second-order DSM, the rate reaches 50.11~Mbps, approximately 97\% of the 12-bit reference. A noise model combining the DSM noise-transfer function with matched filtering guides the sampling-rate and signal-band design for effective noise suppression. We also identify a phase-conjugate spectral replica arising from third-order intermodulation between the pilot tone and quantum signal, and evaluate its key-rate penalty using a linear-replica security model. At 3-bit resolution and an oversampling ratio of 24, first-order DSM reduces the experimentally estimated leakage-mode strength by approximately four orders of magnitude. Together, our findings establish the security boundaries and operating conditions of DSM, providing design guidance for low-resolution CVQKD transmitters.

quant-ph↗

arXiv2026

A Fisher-Geometric Approach to Supercell Pair Distribution Function Refinement

Refining atomistic supercells against powder pair distribution functions (PDFs) requires optimizing many coupled atomic coordinates using a much smaller set of radial measurements. Reverse-mode differentiation makes gradients accessible for large systems, but standard first-order optimizers do not explicitly use coupled PDF sensitivities to precondition coordinate updates. We introduce a matrix-free, Fisher-geometric approach that pulls the information metric of the observation likelihood back onto atomic coordinates. The resulting update weights collective displacements according to their PDF response and experimental uncertainty. Under isotropic damping, the update is obtained by solving a damped linear system over $M$ PDF channels, using automatic differentiation to evaluate Jacobian--vector and vector--Jacobian products without constructing the full Jacobian or coordinate curvature matrix. For Gaussian measurement errors, the update recovers classical Levenberg--Marquardt refinement while avoiding dense coordinate-space factorization. The same computational framework accommodates correlated Gaussian errors and robust likelihood-based updates. This approach combines likelihood-aware collective updates with matrix-free backpropagation for direct second-order refinement of large atomistic supercells.

cond-mat.mtrl-sci↗

arXiv2026

VRL-Bench: Benchmarking agents on computer control tasks under finite trial budgets

Learning from trial and error is a promising way to improve language agents on complex tasks such as computer control. Reflexion introduced verbal reinforcement learning, which turns failed trials into text that guides later attempts without updating model parameters. We introduce VRL-Bench, a harness for fair evaluation of trial-and-error learning under finite trial budgets. Across three models on MiniWoB and WebShop, we evaluate updates from several prominent verbal-memory methods spanning Reflexion and later work: each improves observed success over memory-free retry in some settings but reduces it in others. Replay experiments show that using reflection can reduce success rates, revealing a trade-off between exploiting experience and continued exploration. We propose VEX$^2$, a verbal exploration--exploitation scheduler that uses a language model to jointly select policies and allocate the remaining trial budget. VEX$^2$ is the only evaluated update to achieve positive observed success-rate gains over retry in all six settings.

cs.AI↗

arXiv2026

Hieronym: Leveraging Hierarchical Multi-Source Information for Function Renaming in Stripped Binary

Function renaming in stripped binaries can substantially assist reverse engineers by improving code readability, yet it is a challenging task. The difficulty stems from the need to accurately capture function semantics from low-level binary code across diverse instruction sets, architectures, and compiler optimizations, and to express these semantics in concise, human-readable names. Existing approaches either inadequately capture comprehensive function semantics or exhibit limited generalization to previously unseen binaries. In this paper, we present Hieronym, a generative large language model (LLM)-based framework for stripped binary function renaming. Hieronym adopts a hierarchical summarization-driven domain adaptation strategy and integrates multi-source information, including global binary context, local calling context, and intrinsic function semantics, to enhance the LLM's understanding of binary code. To enable systematic evaluation, we further propose a dual-layer evaluation framework that incorporates both token-level and whole-name-level metrics. We evaluate Hieronym on binary functions compiled with four compiler optimization levels (O0-O3) for four architectures (x64, x86, ARM, and MIPS). Experimental results demonstrate that Hieronym significantly outperforms state-of-the-art methods, achieving token-level improvements of 50.12% in precision, 41.75% in recall, and 45.10% in F1-score, as well as a 79.94% improvement in name-level accuracy, while also exhibiting strong generalization capability. Moreover, experiments on real-world malware samples further validate the practical effectiveness of Hieronym in security-critical scenarios.

cs.SE↗

arXiv2026

Deterministic Minimum-Leakage Continuous-Variable Quantum Key Distribution with Phase-Conjugated Twin Beams

Minimum-leakage continuous-variable quantum key distribution suppresses Eve's Holevo information by engineering the signal ensemble during state preparation. The existing symmetric realization relies on Alice-side heralding. Here we propose an alternative two-mode symmetric realization that is deterministic without heralding. The minimum-leakage condition of the new protocol requires transmitting two phase-conjugated twin beams (PCTB). We show that the PCTB protocol is related to the heralding protocol through a common entanglement-based source but corresponds to a different prepare-and-measure decomposition. The two protocols attain the same secret key rate per transmitted optical mode in the very-large-squeezing limit. For finite squeezing, however, the PCTB protocol requires approximately 3 dB less squeezing to achieve the same performance. We further optimize mode-symmetric Gaussian two-mode entangling-cloner attacks and find that correlated ancillary modes provide only a limited advantage over independent attacks under the minimum-leakage condition. These results establish phase-conjugated twin beams as a deterministic and experimentally appealing route to symmetric minimum-leakage CV-QKD.

quant-ph↗

arXiv2026

The CodeInverter Suite: Structure- and Data-Aware Binary Decompilation with Efficient LLMs

Binary decompilation plays a vital role in various cybersecurity and software engineering tasks. Recently, end-to-end decompilation methods powered by large language models (LLMs) have attracted increasing attention for their ability to generate highly readable source code with minimal human intervention. However, existing LLM-based methods still struggle with reconstructing program structure and logic, achieving accurate data recovery, ensuring data security and privacy, and maintaining computational efficiency. To address these challenges, we propose the CodeInverter Suite, with three main pieces: (1) the CodeInverter Workflow (CIW) is a novel prompt engineering method that incorporates control flow graphs (CFG) and explicit data mappings to enhance structure reconstruction and data recovery during decompilation; (2) building upon CIW, we construct the CodeInverter Dataset (CID), a large-scale domain-specific dataset containing 8.69 million samples enriched with CFGs and data mapping information; (3) we develop CodeInverter Models (CIMs), two lightweight LLMs with 1.3B and 6.7B parameters, enabling efficient inference in privacy-sensitive and resource-constrained environments. Extensive experiments on two benchmark datasets demonstrate that CIW significantly enhances the decompilation performance of various LLMs, with average improvements of 13.07% in re-executability and 23.94% in re-compilability. For our proposed decompilation model, CIM-6.7B achieves state-of-the-art performance in terms of re-executability and readability, outperforming existing LLMs-even with over 100 times more parameters-by an average of 11.03% and 6.27%, respectively.

cs.SE↗

arXiv2026

Dual IMU System for Accurate Gastrointestinal Motility Tracking

Gastrointestinal (GI) motility disorders impair the coordinated movement of food through the digestive tract. The Wireless Motility Capsule was developed to indirectly assess GI motility by recording pressure, pH, and temperature along the GI tract. Although an inertial measurement unit (IMU) enables precise and direct motion tracking, a single-IMU system is highly susceptible to artifacts caused by body movement and respiration. To address this limitation, we propose a dual-IMU system designed to distinguish intrinsic GI motility from extrinsic body motion. One IMU is integrated into an ingestible capsule to monitor internal movement, while a second IMU is worn externally to capture whole-body dynamics. Experimental validation was conducted using a turntable to simulate random body movements and a Stewart platform oscillating at 0.1 Hz to emulate GI motility. Three sensor fusion algorithms, including the Extended Kalman Filter, Madgwick Filter, and Mahony Filter, were benchmarked to calculate the real-time relative orientation between the two IMUs using quaternion analysis. The results show that the dual-IMU system effectively differentiates platform motion from turntable motion, demonstrating its potential to isolate GI-specific motility. Pearson correlation coefficients greater than 0.8 between the baseline and noise-cancelled motility signals indicate strong agreement and effective noise suppression.

eess.SP↗

arXiv2026

MARS-RA: Rank Aggregation for Credit Assignment via Multimodal Comparisons in Embodied Multi-Agent Cooperation

Credit assignment is a fundamental challenge in cooperative multi-agent reinforcement learning, particularly in embodied AI settings characterized by limited and delayed feedback as well as dynamically changing numbers of active agents. We propose MARS-RA, a framework that reformulates credit assignment as a rank aggregation problem using contribution-based pairwise comparisons among agents generated by large multimodal models. This shift from absolute to relative estimation ensures robustness against noise and dynamic agent participation, converting comparison results into contribution scores for potential-based reward shaping. We provide theoretical justification for the convergence and robustness of the proposed framework, and show that Shapley values can be used as an interpretive reference. Experimental results on challenging tasks of different types indicate that MARS-RA can guide agents toward effective cooperation.

cs.AI↗

arXiv2026

Entropy engineering of BF-BT-based high-entropy ceramics for ultra-high energy storage performance

Dielectric capacitors are promising for pulsed power applications, but the energy storage performance of lead-free bulk ceramics is often limited by low breakdown strength and large ferroelectric hysteresis. Herein, a high entropy perovskite oxide BaTi0.2Zr0.2Sn0.2Hf0.2Nb0.1Sc0.1O3 was introduced into the BF-BT matrix to develop lead free high entropy ferroelectric ceramics. Multicomponent B-site substitution induces lattice distortion, enhanced pseudocubic characteristics, relaxor behavior, and grain refinement. These effects suppress polarization hysteresis and electrical conduction, resulting in a significant increase in breakdown strength. A maximum breakdown strength of 840 kV/cm1 and a recoverable energy density of 10.55 J/cm3 were achieved. Entropy-induced microstructural heterogeneity promotes a more uniform electric-field distribution and delays dielectric breakdown. This work demonstrates entropy engineering as an effective route to achieving high breakdown strength and superior energy-storage performance in lead-free ferroelectric ceramics.

cond-mat.mtrl-sci↗

arXiv2026

MinSurf: resolving the atomic-scale stability landscape of mineral surfaces

Mineral surfaces govern interfacial reactivity in carbon mineralization, geo-energy storage, contaminant immobilization, heterogeneous catalysis and electrochemical interface engineering. Yet atomistic simulations often rely on commonly used facets or facet-level stability criteria, while distinct atomic terminations of the same crystallographic orientation are rarely resolved systematically because experimental characterization and density functional theory (DFT) calculations remain costly across large surface spaces. Here we present MinSurf, a high-throughput framework that resolves mineral surface selection as a surface-energy and morphology problem. MinSurf integrates surface enumeration, DFT labelling, machine-learning interatomic potentials and Wulff construction to predict stable terminations, surface-energy landscapes and equilibrium crystal morphologies. Applied to ten representative minerals, MinSurfSet comprises 764 surface slabs, with 90 corresponding oriented unit cells constructed as bulk references for surface-energy evaluation. The resulting MinNEP model predicts DFT surface energies with a mean absolute error of 0.0119 eV per Angstrom squared and achieves an overall acceleration of 1.14 x 10^4 relative to DFT. MinNEP preserves the DFT-derived morphology-determining surface-energy hierarchy and reproduces the dominant Wulff-exposed facets, while X-ray diffraction provides an independent crystallographic consistency check for alpha-quartz benchmark. By linking atomic terminations, surface energies and equilibrium morphologies, MinSurf provides reproducible and physically representative surface models for high-throughput simulations of mineral interfaces across energy, environmental and advanced inorganic materials.

cond-mat.mtrl-sci↗

arXiv2026

TowerMind: A Tower Defence Game Learning Environment and Benchmark for LLM as Agents

Recent breakthroughs in Large Language Models (LLMs) have positioned them as a promising paradigm for agents, with long-term planning and decision-making emerging as core general-purpose capabilities for adapting to diverse scenarios and tasks. Real-time strategy (RTS) games serve as an ideal testbed for evaluating these two capabilities, as their inherent gameplay requires both macro-level strategic planning and micro-level tactical adaptation and action execution. Existing RTS game-based environments either suffer from relatively high computational demands or lack support for textual observations, which has constrained the use of RTS games for LLM evaluation. Motivated by this, we present TowerMind, a novel environment grounded in the tower defense (TD) subgenre of RTS games. TowerMind preserves the key evaluation strengths of RTS games for assessing LLMs, while featuring low computational demands and a multimodal observation space, including pixel-based, textual, and structured game-state representations. In addition, TowerMind supports the evaluation of model hallucination and provides a high degree of customizability. We design five benchmark levels to evaluate several widely used LLMs under different multimodal input settings. The results reveal a clear performance gap between LLMs and human experts across both capability and hallucination dimensions. The experiments further highlight key limitations in LLM behavior, such as inadequate planning validation, a lack of multifinality in decision-making, and inefficient action use. We also evaluate two classic reinforcement learning algorithms: Ape-X DQN and PPO. By offering a lightweight and multimodal design, TowerMind complements the existing RTS game-based environment landscape and introduces a new benchmark for the AI agent field. The source code is publicly available on GitHub(https://github.com/tb6147877/TowerMind).

cs.AI↗

arXiv2026

Divide-and-Conquer Inference for Large-Scale Visual Recognition with Multimodal Large Language Models

Multimodal Large Language Models (MLLMs) have demonstrated strong capabilities across a wide range of vision language tasks. However, when applied to large scale image classification, their performance degrades significantly as the label space expands a phenomenon we define as Performance Collapse in Long Sequence Recognition. Through an information theoretic analysis, we reveal that this collapse stems from a fundamental conflict between the escalating information entropy and the prominent attention dilution and decay within attention mechanisms, which impairs the model's ability to maintain a sufficient signal-to-noise ratio when processing extremely long prompts. To mitigate this, we propose Divide-and-Conquer Inference (DCI), a novel test-time scaling strategy for visual recognition with MLLMs. DCI recursively decomposes complex global classification tasks into multiple simpler, localized subproblems and employs a dynamic pruning mechanism to compress the search space. This method effectively improves the local signal to noise ratio and model accuracy by mitigating the inherent weight dilution issues in long-sequence inference. Moreover, while traditional self-attention incurs a prohibitive quadratic computational complexity, DCI achieves more favorable scaling behavior and substantially accelerates inference in large scale classification scenarios. Extensive experiments on benchmarks such as ImageNet-1K and ImageNet-21K demonstrate that DCI consistently improves classification accuracy. This enables lightweight open-source models to rival or even surpass frontier closed-source giants without any additional training or fine-tuning. As a model-agnostic, plug-and-play paradigm, DCI offers an efficient approach for scaling the inferential precision of MLLMs in large-scale scenarios.

cs.CV↗

arXiv2026

Domain-Wall Mediated Polarization Switching in Ferroelectric AlScN: Strain Relief and Field-Dependent Dynamics

While scandium-doped aluminum nitride (AlScN) exhibits robust ferroelectricity and excellent thermal stability, its utility is limited by an exceptionally high coercive field ($E_c$) for polarization switching. Unraveling the atomistic switching dynamics is therefore critical for tailoring $E_c$. Here, we combine density functional theory and machine-learning molecular dynamics to elucidate the polarization switching mechanisms in AlScN over various Sc concentrations and applied electric fields. We find that excessive lattice strain strictly prohibits collective polarization switching, but the pre-existing domain walls relieve strain and lead to a distinct switching dynamics -- dictating a field-dependent switching mechanism. At low electric fields, switching occurs via gradual domain-wall propagation consistent with the Kolmogorov-Avrami-Ishibashi model. In contrast, high fields stimulate additional nucleation, driving a rapid, homogeneous reversal process described by the simultaneous non-linear nucleation and growth model. These findings highlight the critical role of domain-wall dynamics and suggest domain engineering as a viable strategy to tailor coercive fields in AlScN and related ferroelectrics.

cond-mat.mtrl-sci↗

arXiv2026

Kirigami-Structured Electronic Capsule for Long-Term Continuous Gastric Monitoring

Ingestible electronic systems enable non-invasive, in situ sensing within the gastrointestinal (GI) tract, yet clinical translation has been limited by uncontrolled transit, short operational lifetimes, and unreliable wireless communication that prevent continuous monitoring. Here, we present a gastric-resident ingestible robotic platform that achieves week-long operation through integration of a bioinspired, electrically triggered release mechanism with a kirigami-enabled electronic architecture. A kirigami-patterned flexible printed circuit board spans the capsule body and deployable superelastic arms, enabling high-density integration of sensing, power management, and wireless modules within a constrained volume while tolerating large mechanical deformation during gastric residence. Stable retention and on-demand disassembly are achieved using thermally responsive polycaprolactone joints that transition from rigid to compliant states under electrical activation, avoiding dependence on variable chemical triggers. Reliable telemetry in the highly attenuating gastric environment is maintained using a dual-band Bluetooth Low Energy and sub-gigahertz module with RSSI- and throughput-aware adaptive transmission, balancing link robustness and energy consumption. We demonstrate long-term, continuous monitoring of gastric radiation exposure, enabling early detection of dose accumulation and providing a promising in vivo alternative to wearable or handheld dosimeters. Swine studies confirm stable gastric residence, sustained real-time telemetry, and safe gastrointestinal passage following triggered disassembly. This work establishes kirigami-enabled integration as a scalable strategy for long-term gastric-resident robotic systems.

eess.SY↗

arXiv2026

Step-level Denoising-time Diffusion Alignment with Multiple Objectives

Reinforcement learning (RL) has emerged as a powerful tool for aligning diffusion models with human preferences, typically by optimizing a single reward function under a KL regularization constraint. In practice, however, human preferences are inherently pluralistic, and aligned models must balance multiple downstream objectives, such as aesthetic quality and text-image consistency. Existing multi-objective approaches either rely on costly multi-objective RL fine-tuning or on fusing separately aligned models at denoising time, but they generally require access to reward values (or their gradients) and/or introduce approximation error in the resulting denoising objectives. In this paper, we revisit the problem of RL fine-tuning for diffusion models and address the intractability of identifying the optimal policy by introducing a step-level RL formulation. Building on this, we further propose Multi-objective Step-level Denoising-time Diffusion Alignment (MSDDA), a retraining-free framework for aligning diffusion models with multiple objectives, obtaining the optimal reverse denoising distribution in closed form, with mean and variance expressed directly in terms of single-objective base models. We prove that this denoising-time objective is exactly equivalent to the step-level RL fine-tuning, introducing no approximation error. Moreover, we provide numerical results, which indicate our method outperforms existing denoising-time approaches.

cs.LG↗