arXiv subjects

Qingwen Liu

Publications and source records attributed to Qingwen Liu.

At least 19 recordsLinked to original sources

arXiv2026

Odometry-Aided Real-Time Mapping for Underwater Robots Using Forward-Looking Sonar

Reliable perception is essential for underwater vehicles operating in complex environments, where light attenuation and scattering often degrade visibility and compromise optical sensing. Forward-looking sonar (FLS) offers an alternative by providing high-frame-rate acoustic imaging under poor optical conditions. However, real-time FLS mapping remains challenging due to unresolved target elevation, spatially non-uniform noise, and fragmented target boundaries, which hinder feature extraction and introduce geometric ambiguity during projection. To address these challenges, we propose a cascaded feature reconstruction pipeline combining fast Fourier transform (FFT)-based denoising, fast multiscale constant false alarm rate (MCFAR) detection, and gradient-adaptive boundary connection to extract geometric features from degraded sonar images with low latency. We integrate attitude-aware geometric projection with incremental occupancy accumulation to construct a depth-referenced 2.5D map for local mapping in confined underwater environments. The sonar's vertical position is referenced to an external sensor, while target elevation is assigned under an explicit geometric assumption rather than measured directly by FLS. Experiments in a 3 m X 5 m pool demonstrate centimeter-scale planar mapping accuracy, with a root-mean-square error (RMSE) below 3 cm across three sequences and an average processing time of 42.4 ms per frame.

cs.RO↗

arXiv2026

Integrated Sensing and Communication System Based on Radio Frequency Resonance Beam

To address the complex beam control in traditional multiple-input multiple-output (MIMO) systems, researchers have proposed adaptive beam alignment using retro-directive antenna (RDA) arrays. This approach creates echo resonance between the base station (BS) and user equipment (UE), significantly reducing computational load. However, conventional resonant beam systems (RBS) suffer from echo interference due to the shared uplink and downlink frequency. Therefore, this paper proposes an innovative resonance beam-based integrated sensing and communication (RB-ISAC) system designed for efficient passive sensing and bidirectional communication. In this system, the UE operates passively, with both the BS and UE utilizing a phase conjugation and frequency conversion structure to decouple uplink and downlink carrier frequencies, ensuring continuous electromagnetic wave oscillation between the two ends. Effective compensation for signal propagation loss enables resonance after multiple oscillations. At this point, the beam's field forms a low-diffraction-loss, highly focused pattern, automatically aligning the transmitter and receiver. This enables high-precision passive positioning alongside robust uplink and downlink communication. Simulation results demonstrate the proposed system achieves resonance within multiple iterations, supporting uplink and downlink communication up to 5 m, and enabling passive direction of arrival (DOA) estimation with an error under 2$^\circ$ .

eess.SP↗

arXiv2026

FoV and Efficiency Optimization for Resonant Beam SLIPT with Telescope Integration

Meeting the large bandwidth demands of wireless communication for mobile Internet of Things (IoT) devices while enhancing their endurance is a significant challenge. Simultaneous lightwave information and power transfer (SLIPT) technology offers the potential to realize wireless charging and signal transfer, making it suitable for supporting autonomous vehicles and drones. The resonant beam system (RBS) leverages the self-aligning property of a spatially distributed laser resonator (SSLR), allowing energy transmission from the transmitter to the receiver without mechanical alignment. However, the existing resonant beam SLIPT system exhibits a limited field of view (FoV) and transmission efficiency, facing challenges in practical applications. In this paper, we propose a resonant beam SLIPT system enhanced by incorporating an internal telescope and optimizing the communication, energy transfer, and FoV performance by solving the Pareto front set of the system's achievable performance region. The results indicate that the optimized FoV is increased by $17\%$, reaching $\pm26.8^\circ$, while its average end-to-end efficiency is improved by $145\%$, achieving $5.4\%$.

physics.optics↗

arXiv2026

World Model Control by Trajectory Reachability Metrics

Latent world models can learn representations that contain information needed for control, while the downstream controller may still rank candidate actions poorly when it relies on terminal latent distance alone. We study this failure in a fixed encoder and introduce trajectory reachability metrics (TRM), a small temporal pairwise cost trained from logged trajectories and used to rank predicted endpoints of a candidate action sequence against a goal. In the TwoRoom evaluation on 100 episodes from a high distance range, the original controller with LeWorldModel reaches 7.0% mean success. Temporal TRM trained after excluding all evaluation episodes reaches 96.7%, shuffled label controls stay at 0.0%, and TRM also improves PLDM from 32.7% to 84.0%. TopoNav, a separate pixel navigation task, shows the same link between repaired ranking and closed loop control. The selection audit with shared candidates (SASC) and rowspace interventions show that the gain comes from reweighting latent directions carrying the action decision. The XY rowspace contributes less than 1% of terminal latent MSE but carries most of the information needed for control. Coverage and boundary tests define the scope. Balanced doorway coverage restores a 0.0% coverage failure to 100.0% success, while an unseen wall orientation and contact rich PushT expose layout, dynamics, and recovery limits.

cs.LG↗

arXiv2026

Action from Adjacent Set in Physical Space Outperforms the Best Prediction in World Models

Controllers based on sampling and latent world models assign a predicted terminal cost to each candidate action sequence, choose the minimum, execute its first action block, and replan. This rule can fail even when the terminal cost perfectly and accurately reflects the true task objective in the physical world. Residual prediction error can give an infeasible sequence an anomalously low cost, and a larger proposal pool gives such errors more chances to outrank feasible alternatives. We call this conditional failure proposal overgeneration. In Cube candidate execution audits, increasing the total proposal budget from 72 to 288 reduces the feasibility of selection by minimum latent cost from .375 to .062 for position targets and from .344 to .031 for targets defined by position and yaw, although every larger pool contains a feasible sequence. We introduce Adjacent Set Action Reconstruction (ASAR). Among proposals with low cost, ASAR identifies an adjacent set using standardized early action prefixes and reconstructs a full action sequence through locally weighted aggregation with a light anchor from the sequence with minimum cost. On a Carry and Release evaluation set of 75 queries, Kernel ASAR improves event completion success over matching selection by 28.0, 24.0, and 18.7 percentage points under latent cost and by 18.7, 20.0, and 17.3 points under a trajectory reachability cost at 72, 144, and 288 proposals. Analysis of finite proposal pools characterizes selection risk from the lower tail, separation by a related radius support statistic, and sequence containment under an explicit local feasibility condition.

cs.RO↗

arXiv2026

MIRA: Medical Image Reflection for Agentic Diagnosis

Medical visual agents can use tools to inspect images and retrieve external knowledge, but indiscriminate tool use may introduce noisy or misleading evidence. Reliable diagnosis therefore requires not only acquiring additional observations, but also verifying whether tool actions are necessary and whether the resulting evidence supports the current hypothesis. We introduce MIRA (Medical Image Reflection for Agentic Diagnosis), a medical visual diagnostic framework for autonomous evidence search and reflective verification. MIRA dynamically invokes image-processing operations, including zooming, grounding, pointing, rotation, and measurement, as well as web search, while evaluating the relevance and consistency of the acquired evidence. We develop MIRA through a two-stage training strategy. First, a tool-augmented Monte Carlo Tree Search data engine explores diverse diagnostic hypotheses and jointly verifies visual grounding accuracy and semantic consistency to construct supervised fine-tuning trajectories. Second, reinforcement learning further improves decision-making through online reflective principle evolution: failure cases are distilled into candidate principles, and only principles that improve held-out rollout rewards are retained. Across nine medical visual reasoning benchmarks, MIRA achieves an average score of 64.73, improving its Qwen3-VL-8B backbone by 7.44 points. It also increases useful tool-use judgments from 56.2% to 73.8% and reduces harmful judgments from 8.9% to 1.6%. Qualitative analyses show that MIRA can re-examine evidence, correct premature conclusions, and adapt its tool-use strategy. Project page: https://MIRA-VL.github.io/

cs.CV↗

arXiv2026

Layered Architecture for Mobile Intelligence

Artificial intelligence (AI) is rapidly evolving from a centralized computing capability into a pervasive infrastructure that interacts directly with the physical world. While recent perspectives highlight the roles of energy, chips, infrastructure, models, and applications in enabling large-scale AI systems, these frameworks primarily assume a static, cloud-centric computing paradigm. However, emerging intelligent applications, including autonomous vehicles, drones, robots, and wearable systems, require AI to operate in highly dynamic and mobile environments. This shift introduces mobility as a fundamental constraint across the entire AI ecosystem, affecting energy supply, computation, and intelligence deployment. In this article, we introduce the concept of the Mobile AI Stack, a mobility-aware architectural framework that integrates five tightly coupled layers: mobile energy networks, energy-efficient AI chips, cloud-edge-mobile infrastructure, distributed AI models, and embodied AI applications. The proposed framework provides a systematic perspective for understanding how energy delivery, computing architectures, communication networks, and AI algorithms must co-evolve to support large-scale mobile intelligence. We further discuss key research challenges and future directions toward building scalable, reliable, and energy-efficient mobile AI systems. Mobile AI Stack offers a conceptual blueprint of the next-generation infrastructure which deeply integrates the networks of computation, energy, and communications for mobile intelligence.

cs.NI↗

arXiv2026

AB-Sync: Attention-Based Slot-Level Clock Synchronization Method for UWB-TDOA Localization Networks

Ultra-wideband (UWB) time-difference-of-arrival (TDOA) localization networks provide high-update-rate indoor location services for IoT and cyber-physical applications, but their accuracy depends on nanosecond-level clock synchronization among anchors. Existing wireless clock synchronization (WCS) methods typically estimate clock states at the synchronization-stage or interval level, whereas TDMA-based UWB-TDOA systems localize tags from blinks transmitted in discrete short slots inside each synchronization stage. We identify this granularity mismatch as a source of residual TDOA error and present AB-Sync, an attention-based slot-level clock synchronization method. AB-Sync models the relationship between the slot-specific clock-speed ratio required by a target tag blink and neighboring clock-fluctuation observations, thereby enabling tag-slot-level timestamp mapping without adding extra UWB synchronization messages. On a real UWB-TDOA testbed, AB-Sync reduces the multi-anchor average TDOA ranging STD.V by 9.4% and improves representative static localization accuracy by 18.6% compared with Deferred+3S-KF, the leading low-overhead baseline in our evaluation. In a five-slot multi-tag experiment, AB-Sync consistently improves localization stability across all TDMA slots, reducing STD.V by 5.3% on average and up to 16.2% per slot with no extra UWB synchronization overhead.

cs.NI↗

arXiv2026

Beyond Mode Collapse: Distribution Matching for Diverse Reasoning

On-policy reinforcement learning methods like GRPO suffer from mode collapse: they exhibit reduced solution diversity, concentrating probability mass on a single solution once discovered and ceasing exploration of alternative strategies. We show this stems from reverse KL minimization's mode-seeking behavior, which reinforces the first high-reward trajectory found rather than maintaining a distribution over multiple diverse solutions. We propose DMPO (Distribution-Matching Policy Optimization), which prevents mode collapse through principled approximation of forward KL minimization. DMPO constructs a group level target distribution over sampled trajectories proportional to their rewards, then aligns the policy distribution to this target. This provides mode-covering behavior without requiring sampling from the intractable global target distribution, enabling sustained exploration throughout training. We validate DMPO on NP-hard combinatorial optimization, where exponentially many feasible solutions exist but only a few approach optimality, an ideal testbed for evaluating exploration. DMPO achieves 43.9% Quality Ratio on text-based NP-Bench (vs. GRPO's 40.1%) and 43.1% on vision-based NP-Bench (vs. 38.4%), demonstrating 9% and 12% relative improvements respectively. These gains generalize to mathematical reasoning (+2.0%) and out-of-domain tasks (+2.3%), showing that diversity-preserving training enhances general reasoning capabilities across modalities. Our work establishes distribution matching as a practical, principled approach to preventing mode collapse in on-policy RL, with consistent quality improvements demonstrating sustained exploration across diverse reasoning tasks.

cs.AI↗

arXiv2026

OPT-BENCH: Evaluating the Iterative Self-Optimization of LLM Agents in Large-Scale Search Spaces

Large Language Models (LLMs) have demonstrated remarkable capabilities in reasoning and tool use. However, the fundamental cognitive faculties essential for problem solving, including perception, reasoning, and memory, remain the stable core of intelligence. Unlike memorizing specific patterns, humans succeed in novel environments by applying these intrinsic faculties to adapt and optimize. Yet, whether LLMs possess this essential capacity, namely the ability to continuously refine solutions in response to dynamic environmental feedback, remains underexplored. To address this challenge, we introduce OPT-BENCH, a benchmark for evaluating self-improvement capabilities in large-scale search spaces. By combining 20 machine learning tasks with 10 classic NP-hard problems, OPT-BENCH provides a rigorous setting to assess whether agents can adapt through intrinsic self-reflection rather than rote tool application. We further propose OPT-Agent, a framework that emulates human-like cognitive adaptation. It operates through a general perception, memory, and reasoning loop, iteratively refining solutions based on environmental feedback. Through extensive experiments on 19 LLMs from 7 model families, including reasoning models, general models, and open-source models ranging from 3B to 235B parameters, we demonstrate that stronger models are more effective at leveraging feedback signals for self-improvement. However, this upper-bound adaptability remains fundamentally constrained by the models' base capacity, and even the most advanced LLMs still fall short of human expert performance.

cs.AI↗

arXiv2026

Forge: Quality-Aware Reinforcement Learning for NP-Hard Optimization in LLMs

Large Language Models (LLMs) have achieved remarkable success on reasoning benchmarks through Reinforcement Learning with Verifiable Rewards (RLVR), excelling at tasks such as math, coding, logic, and puzzles. However, existing benchmarks evaluate only correctness, while overlooking optimality, namely the ability to find the best solutions under constraints. We propose OPT-BENCH, the first comprehensive framework for training and evaluating LLMs on NP-hard optimization problems through quality-aware RLVR. OPT-BENCH provides three key components: a scalable training infrastructure with instance generators, quality verifiers, and optimal baselines across 10 tasks; a rigorous benchmark with 1,000 instances evaluating both feasibility, measured by Success Rate, and quality, measured by Quality Ratio; and quality-aware rewards that enable continuous improvement beyond binary correctness. Training on Qwen2.5-7B-Instruct-1M with 15K examples achieves 93.1% SR and 46.6% QR, significantly outperforming GPT-4o, which achieves 29.6% SR and 14.6% QR. Beyond optimization, training on OPT-BENCH transfers to diverse tasks, including mathematics (+2.2%), logic (+1.2%), knowledge (+4.1%), and instruction following (+6.1%). Our analysis reveals that quality-aware rewards improve solutions by 28.8% over binary rewards, and that task diversity drives generalization more than data quantity, offering insights into RLVR scaling for complex reasoning.

cs.AI↗

arXiv2026

Escaping the Context Bottleneck: Active Context Curation for LLM Agents via Reinforcement Learning

Large Language Models (LLMs) struggle with long-horizon tasks due to the "context bottleneck" and the "lost-in-the-middle" phenomenon, where accumulated noise from verbose environments degrades reasoning over multi-turn interactions. To address this issue, we introduce a symbiotic framework that decouples context management from task execution. Our architecture pairs a lightweight, specialized policy model, ContextCurator, with a powerful frozen foundation model, TaskExecutor. Trained via reinforcement learning, ContextCurator actively reduces information entropy in the working memory. It aggressively prunes environmental noise while preserving reasoning anchors, that is, sparse data points that are critical for future deductions. On WebArena, our framework improves the success rate of Gemini-3.0-flash from 36.4% to 41.2% while reducing token consumption by 8.8% (from 47.4K to 43.3K). On DeepSearch, it achieves a 57.1% success rate, compared with 53.9%, while reducing token consumption by a factor of 8. Remarkably, a 7B ContextCurator matches the context management performance of GPT-4o, providing a scalable and computationally efficient paradigm for autonomous long-horizon agents.

cs.AI↗

arXiv2026

COINBench: Moving Beyond Individual Perspectives to Collective Intent Understanding

Understanding human intent is a high-level cognitive challenge for Large Language Models (LLMs), requiring sophisticated reasoning over noisy, conflicting, and non-linear discourse. While LLMs excel at following individual instructions, their ability to distill Collective Intent - the process of extracting consensus, resolving contradictions, and inferring latent trends from multi-source public discussions - remains largely unexplored. To bridge this gap, we introduce COIN-BENCH, a dynamic, real-world, live-updating benchmark specifically designed to evaluate LLMs on collective intent understanding within the consumer domain. Unlike traditional benchmarks that focus on transactional outcomes, COIN-BENCH operationalizes intent as a hierarchical cognitive structure, ranging from explicit scenarios to deep causal reasoning. We implement a robust evaluation pipeline that combines a rule-based method with an LLM-as-the-Judge approach. This framework incorporates COIN-TREE for hierarchical cognitive structuring and retrieval-augmented verification (COIN-RAG) to ensure expert-level precision in analyzing raw, collective human discussions. An extensive evaluation of 20 state-of-the-art LLMs across four dimensions - depth, breadth, informativeness, and correctness - reveals that while current models can handle surface-level aggregation, they still struggle with the analytical depth required for complex intent synthesis. COIN-BENCH establishes a new standard for advancing LLMs from passive instruction followers to expert-level analytical agents capable of deciphering the collective voice of the real world. See our project page on COIN-BENCH.

cs.IR↗

arXiv2026

Simultaneous Self-Localization and Base Station Localization with Resonant Beam

High-precision positioning in GPS-denied environments is a demanding but challenging technology. Resonant Beam Positioning (RBP) utilizes a resonant beam with properties such as energy focusing, self-establishment, self-alignment, and passive operation, offering a promising solution for this task. However, traditional RBP algorithms require a fixed number of resonant beam base stations, which can be costly to expand coverage. To address this limitation, we propose a distributed resonant beam positioning (DRBP) system that simultaneously estimates the base station and mobile target (MT) positions. Firstly, the MT receives resonant beam samples to locate the base station in the limited field of view (FoV) region. Subsequently, it estimates self-position based on the known locations of the base stations. During moving, the DRBP system facilitates self-positioning on the MT side, enabling dynamic expansion of both the number of base stations and the coverage area. Numerical results demonstrate that DRBP achieves a positioning root mean square error (RMSE) of $0.1$ m and a rotation RMSE of 2$^\circ$, validating the system's high accuracy.

physics.optics↗

arXiv2026

High-Resolution Multi-Target DOA Estimation for Resonant Beam Systems

Direction of arrival (DOA) estimation technology offers a promising solution to address the sensing and positioning demands of Internet of Things (IoT) devices. Optical resonant beam systems (RBS), owing to their inherent characteristics of self-alignment, self-established energy focusing, and passive target sensing, make them naturally suited for {\color{blue}DOA} estimation in IoT scenarios. However, RBS suffer from limited angular resolution and a narrow field of view (FoV) in multi-target environments. To overcome these limitations, this paper proposes a high-resolution wide-field-of-view resonant beam DOA estimation system (RB-HWDOA). The RB-HWDOA integrates an optical spectrum-based DOA estimation algorithm (OSB-DOA), which leverages amplitude information in the two-dimensional Fourier spectrum of the resonant beam, {\color{blue}overcoming the resolution limit imposed by the beam size in spatial-domain methods}. Furthermore, we designed a {\color{blue}telescope} modulation (TM) structure to correct phase and direction mismatches, enabling a multi-Tx framework that focuses beams onto a common sensing module, thereby extending the effective FoV. Combined with the OSB-DOA algorithm, this design supports high-resolution DOA estimation for {\color{blue}multiple targets simultaneously over a wide FoV}. Simulation results show that OSB-DOA resolves angular separations down to $0.1^{\circ}$ across multiple resonant beams, remains robust under noise, {\color{blue}and the TM architecture enables multi-Tx integration for wide-FoV coverage}, making RB-HWDOA a scalable and efficient solution for passive multi-target DOA estimation in complex IoT environments.

physics.optics↗

arXiv2026

On the Stability of Spatially Distributed Cavity Laser and Boundary of Resonant Beam SLIPT

Spatially distributed cavity (SDC) lasers are a promising technology for simultaneous light information and power transfer (SLIPT), offering benefits such as increased mobility and intrinsic safety, which are advantageous for various Internet of Things (IoT) devices. \mll However, achieving beam transmission over meter-level long working distances presents significant challenges from cavity stability constraints, manufacturing/assembly tolerances, and diffraction losses\mrr. This paper conducts a theoretical investigation of the fundamental restrictions limiting long-range resonant beam generation. We investigate cavity stability and beam characteristics, and propose a binary-search-based Monte Carlo simulation algorithm as well as a linear approximation algorithm to quantify the maximum acceptable tolerances for stable operation. \mll Numerical results indicate that the stable region contracts sharply as distance increases. For fixed-component systems, an acceptable tolerance of 0.01 mm restricts the achievable transmission distance to less than 2 m. \mrr To address this limitation, we also prove the feasibility of long-range beam formation using precision adjustable elements, paving the way for advanced engineering applications. \mll Experimental results verified this assumption, demonstrating that by tuning the stable region during assembly, the transmission distance could be extended to 2.8 m. \mrr This work provides essential theoretical insights and practical design guidelines for realizing stable, long-range SDC systems.

eess.SY↗

arXiv2025

Self-Alignment Resonant Beam Empowers Beamforming without Estimation and Control for 6G IoT

The integration of communication, sensing, and wireless power transfer (WPT) is a cornerstone of 6G intelligent IoT. However, relying on traditional beamforming imposes prohibitive overheads due to complex channel state information (CSI) estimation and active beam scanning, particularly in dynamic environments. This paper presents a comprehensive review of the radio frequency resonant beam system (RF-RBS), a native physical-layer paradigm that circumvents these limitations. By deploying retro-directive antenna arrays (RAA) at transceivers, RF-RBS establishes a self-sustaining cyclic electromagnetic loop. This mechanism inherently enables self-aligning, high-gain beamforming through positive feedback, eliminating the reliance on digital CSI processing. We analyze the system's architecture and its capability to support high-efficiency WPT, robust communication, and millimeter-level passive positioning. Finally, we evaluate the implementation challenges and strategic value of RF-RBS in latency-sensitive 6G scenarios, including unmanned systems and industrial automation.

eess.SP↗

arXiv2025

Design and Analysis of Phase Conjugation-Based Self-Alignment Beamforming for RIS-Assisted Terahertz SWIPT

Terahertz (THz) simultaneous wireless information and power transfer (SWIPT) is a promising technology for enabling ultra-high-rate and low-latency communications in massive battery-free Internet of Things (IoT) deployments for 6G networks. However, conventional THz systems rely on narrow directional beams that necessitate precise alignment, typically achieved through high-overhead beam scanning procedures, which fundamentally at odds with the energy constraints of battery-free IoT devices. In this paper, we propose a novel self-alignment architecture for THz SWIPT leveraging a reconfigurable intelligent surface (RIS) to eliminate complex beam scanning. By integrating phase conjugate circuits at both the base station and user equipment, the RIS facilitates a resonance-based bidirectional retro-reflection mechanism, enabling the system to autonomously converge to an aligned state without manual intervention. We develop an analytical channel transfer model and a power cycle model to characterize the resonance-assisted beam alignment process and power transfer efficiency. Simulation results demonstrate that the RIS-enabled system achieves effective spatial power concentration with significant sidelobe suppression, leading to a communication capacity of 127.84 Gbit/s and a received power of 13.62 mW over a 2.2-meter link.

eess.SP↗