Searcharxiv⌕ Search

arXiv subjects

Yijun Chen

Publications and source records attributed to Yijun Chen.

At least 19 recordsLinked to original sources
arXiv2026

Benchmarking Regularization Methods For 3D Radio Tomographic Imaging

Integrated Sensing and Communication (ISAC) is an important technology for 6G, enabling wireless systems to perceive their physical environment. Radio Tomographic Imaging (RTI) is a potential technique for device-free sensing in ISAC, reconstructing object locations and shapes from Received Signal Strength (RSS) measurements. It can complement other techniques, such as radar and LiDAR, by providing a narrowband modality with wavelength resolution. Recently, an extended version of RTI, known as the extended phaseless Rytov approximation (x3DPRA), has been developed to enhance RTI reconstruction quality, estimate material parameters, and extend RTI from two to three dimensions (3D). However, the 3D formulation is even more ill-posed than the 2D form, as the number of measurements is far fewer than the number of unknown voxels. In this work, we systematically evaluate and compare three distinct regularization approaches (Ridge, Total Variation, and Tensor Nuclear) within the x3DPRA framework. We provide a detailed performance analysis by evaluating both reconstruction quality and computational runtime across the three methods. Our findings show that Total Variation regularization provides the best reconstruction quality but not the fastest runtime, providing guidance for developing high-resolution 3D RTI systems for future ISAC applications.

eess.IV↗
arXiv2026

EM^2Mem: Event-Centric Multimodal Memory for Large Language Models

Multimodal memory offers a scalable interface for long-video question answering, but existing methods often retrieve captions, frames, transcripts, summaries, or graph facts as isolated fragments. Although searchable, such fragments are not generation-ready: language models must reconstruct cross-modal and temporal alignments at inference time, when context is limited and attribution is difficult. We propose EM^2Mem, an event-centric multimodal memory framework that binds heterogeneous evidence to event anchors during memory construction. Each event-indexed memory cell aligns multimodal records, temporal context, graph-linked relations, semantic facts, and provenance, enabling compact evidence readout over grounded multimodal events rather than modality-specific fragments. Across three long-video QA benchmarks, EM^2Mem improves average accuracy over the strongest memory baseline by 2.0, 2.4, and 3.7 points, improves strict event-level Top-5 evidence recall by 7.0 points, and reduces per-query latency by 4.67 times and total inference tokens by 63.66% (The code will be integrated into https://github.com/zjunlp/LightMem).

cs.CL↗
arXiv2026

Learning to Evaluate Before Improving: Automatic Rubric Induction for Automatic Research Agents

Autonomous scientific research agents are increasingly applied to end-to-end scientific workflows, including literature review, data analysis, experimentation, and report generation. However, open-ended research tasks often do not clearly specify the analyses, methods, and success criteria required to complete the task. As a result, agents may miss important analyses, use inappropriate methods, or draw conclusions that are insufficiently supported by evidence. To address the problem, we present AutoSciRub, an evaluation-first framework that induces a task-specific executable rubric before research execution, and uses it to guide execution, criterion-level verification as well as iterative revision. AutoSciRub decomposes an underspecified instruction into atomic scientific goals, grounds them in relevant literature and task-visible data, and synthesizes specific, actionable, and verifiable criteria. The resulting rubric makes implicit experimental and evidential requirements explicit, providing guidance for experiments and analyses. During revision, rubric-guided verification identifies unmet criteria and enables targeted refinement of the research report and its supporting artifacts. On ResearchClawBench, AutoSciRub consistently improves all tested configurations, with an average gain of 2.08 points across three backbone LLMs under the fixed Codex harness and 2.95 points across three agent harnesses using a fixed DeepSeek-V4-Flash backbone. On a randomly sampled 20-task subset of AstaBench E2E Discovery, AutoSciRub further achieves an average improvement of 16.8 points across three agent harnesses, while maintaining or increasing the number of successfully completed tasks. These results demonstrate that evaluation-first guidance provides an effective and generalizable control mechanism for autonomous scientific research (Code: https://github.com/zjunlp/AutoSciRub).

cs.CL↗
arXiv2026

MobileMem: Learning from a Year of Mobile Experiences

The next generation of AI agents is increasingly moving beyond systems that answer isolated questions toward persistent personal assistants that can understand, remember, and continuously learn from users' experiences. Such assistants require long-term memory to accumulate and leverage user-specific experiences over time, yet existing benchmarks remain inadequate for realistic mobile settings, where experiences are heterogeneous, multimodal, evolving, and deeply personal. We introduce MobileMem, a benchmark and framework for studying on-device long-term memory, grounded in a year-scale collection of mobile experiences. MobileMem employs a knowledge-grounded synthesis pipeline to construct coherent and temporally consistent long-horizon trajectories from user-app sessions. It provides complementary text and multimodal settings covering multi-hop and temporal reasoning, knowledge updating, and implicit preference inference. Specifically, MobileMem enables agents to remember the past, understand the present, and adapt to the future. By modeling experiences rather than isolated facts, MobileMem moves memory beyond information retrieval toward experiential intelligence for continuous personal learning.

cs.AI↗
arXiv2026

EviBack: Search-Agent Reinforcement Learning via Evidence-Constrained Teacher Backoff

Reinforcement learning enables Agentic RAG systems to learn multi-turn search from verifiable outcome rewards, but all- zero rollout groups provide no comparative signal and may hide useful search behavior. We present EviBack, an evidence- constrained Teacher backoff that supplies auxiliary super- vision to such groups while preserving verifiable Actor re- wards. It separates evidence assessment from answer refine- ment, preventing reference answers from overriding evidence- insufficiency judgments. A fully automated, end-to-end GPT- 5.5-assisted APE pipeline starts from a manually authored single-prompt dual-task Teacher, automatically partitions and labels rollout data, and performs ablation, task decomposition, evaluation, and selection to produce a gated two-stage Teacher. Compared with the manual design, the resulting Teacher im- proves downstream F1 and valid-answer rate while reduc- ing search, duplicate queries, and forced termination. Across seven open-domain QA benchmarks and three Qwen3 scales, EviBack improves F1 over Search-R1 and raises both single- and multi-hop macro F1. We guarantee that the code will be made publicly available at a later stage.

cs.AI↗
arXiv2026

LightMem-Ego: Your AI Memory for Everyday Life

Personal AI assistants on mobile and wearable devices continuously perceive users' daily lives through visual and audio streams. However, answering queries about past experiences requires lightweight multimodal memory that can continuously accumulate, organize, and retrieve long-term experiences, which remains challenging. To address this challenge, we present LightMem-Ego, a lightweight streaming multimodal memory system for everyday-life assistance. The system continuously captures egocentric visual and audio streams, aligns them on a shared timeline, and organizes them into a hierarchical memory consisting of current, short-term, and long-term memory. Given a user query, LightMem-Ego dynamically routes retrieval to the appropriate memory level and generates answers grounded in multimodal evidence. The demonstration can be deployed on smartphones and AI glasses, supporting object finding, conversation recall, life summarization, routine discovery, and personalized assistance. Code is available at https://github.com/zjunlp/LightMem-Ego.

cs.CL↗
arXiv2026

A Dual-Band Reconfigurable Shared-Aperture Antenna Array With Independent Sub-6-GHz and Centimeter-Wave Beam Control

A planar dual-band reconfigurable shared-aperture antenna array is proposed for compact next-generation wireless front ends that require both sub-6-GHz and centimeter-wave (cm-wave) coverage. The array integrates a 2 by 2 sub-6-GHz microstrip dipole array and a 4 by 4 cm-wave stacked patch array within the same aperture, while providing independent beam control in the two bands without conventional T/R modules or beamforming networks. Slot-coupled feeding is employed to separate the radiating aperture from the reconfigurable RF feeding networks and DC bias circuits. PIN-diode-loaded split feeding rings first provide independent 1-bit phase reconfigurability for both bands. A compact reconfigurable $90^{\circ}$ phase shifter is then introduced as an additional phase-control stage, resulting in 2-bit phase control for sub-6 GHz elements and cm-wave subarrays. To reduce cross-band coupling in the compact shared aperture, a double-layer electromagnetic band-gap (EBG) structure is used to suppress cm-wave surface waves and higher-order sub-6-GHz modes excited by the cm-wave elements. A prototype is fabricated and measured. In the sub-6-GHz band, 11 reconfigurable radiation patterns are obtained, including two difference patterns and nine directional beams, with a peak broadside gain of 10.5 dBi. In the cm-wave band, two-dimensional beam scanning up to $\pm40^{\circ}$ is demonstrated with a peak gain of 14.6 dBi in both the E-plane and H-plane. These results show that the proposed architecture can combine dual-band shared-aperture integration and independent reconfigurable beam control in a compact antenna platform.

eess.SP↗
arXiv2026

Fortress and Gatekeeper: Theorizing Transitive Trust in Third-Party Cybersecurity Risk Governance

Third-party vendors, such as analytics platforms, cloud services, identity providers, and software suppliers, are increasingly embedded in digital service delivery. While these arrangements enable scale and specialization, they also move customer data and security-relevant practices into environments that customers rarely see, select, or evaluate. This paper examines this problem through a document analysis of the November 2025 OpenAI-Mixpanel security incident. The incident serves as an illustrative case for showing how a security event in a vendor environment can become a governance and accountability problem for the focal organization that maintains the customer relationship. Drawing on organizational trust research and agency theory, the paper argues that third-party cybersecurity risk is both a trust relationship and a delegation problem. Customers trust the visible service provider, while the provider relies on vendors whose security practices are only partially visible and controllable. The paper develops the concept of transitive trust, where customer trust in a digital service depends on the security practices of vendors authorized by that service provider. It then presents the Fortress and Gatekeeper framework, which explains cybersecurity governance boundaries through trust and data flows rather than formal organizational ownership alone. The analysis develops four propositions concerning vendor integration, metadata exposure, vendor assurance, and data proliferation. The paper contributes to cybersecurity governance scholarship by explaining how delegated data processing creates customer-facing accountability and by identifying implications for vendor tiering, data classification, contractual design, continuous assurance, and data minimization.

cs.CR↗
arXiv2026

TokenPilot: Cache-Efficient Context Management for LLM Agents

As LLM agents are deployed in long-horizon sessions, context accumulation drives up inference costs. Existing approaches utilize text pruning or dynamic memory eviction to minimize token footprints; however, their unconstrained sequence mutations alter layouts, introducing prefix mismatches and cache invalidation. This reveals a critical trade-off between text sparsity and prompt cache continuity. To address this, we present TokenPilot, a dual-granularity context management framework. Globally, Ingestion-Aware Compaction acts as a framework harness to stabilize prompt prefixes and eliminate open-world environmental noise at the ingestion gate. Locally, Lifecycle-Aware Eviction monitors the ongoing residual utility of context segments, enforcing a conservative batch-turn schedule to offload content segments only when task relevance expires. Experiments on PinchBench and Claw-Eval under both isolated and continuous modes demonstrate that TokenPilot reduces costs by 61% and 56% in isolated mode, and 61% and 87% in continuous mode, while maintaining competitive performance compared to prior systems. TokenPilot has been integrated into LightRSI at https://github.com/zjunlp/RSI.

cs.CL↗
arXiv2026

A 3D Formulation of the Extended Phaseless Rytov Approximation

The extended Phaseless Rytov Approximation (xPRA) is a recently proposed device-free RF imaging technique that provides high-resolution reconstructions of the imaging region using only phaseless measurements, such as received signal strength (RSS). Because of its phaseless formulation, it can be implemented straightforwardly using existing wireless commu?nication infrastructure. It also outperforms well-known device?free phaseless RF imaging methods such as Radio Tomographic Imaging (RTI). The linear phaseless formulation used in xPRA(and RTI) makes these methods potentially useful for integrated sensing and communication (ISAC) systems in next generation wireless networks since they do not require wide bandwidths. However, so far, both xPRA and RTI have primarily been formulated in two dimensions (2D). This paper introduces a 3D extension of xPRA, which we call the extended three-dimensional phaseless Rytov approximation (x3DPRA). The novelty of our approach is that it preserves the straightforward implementation advantages of RTI and xPRA while enabling volumetric (3D) imaging. Simulation results show that x3DPRA provides good estimates of location and shape and can also reconstruct object material attenuation. We present the 3D formulation, validate it with a 2D model comparison, and report simulation results demonstrating its performance.

eess.IV↗
arXiv2026

PCDM: A Diffusion-Based Data Poisoning Attack Against Federated Learning Systems

Federated learning (FL) is vulnerable to data poisoning attacks due to its distributed nature. Although recent GAN-based data poisoning methods have indicated the potential of using generative AI to generate seemingly legitimate poisoned data, the inherent consistency of GAN outputs can still reveal a sign of data poisoning. In this paper, we propose a diffusion-based data poisoning framework against FL systems, which leverages a Poisoning-Oriented Conditional Diffusion Model (PCDM) to enable fine-grained control over the local generation of poisoned data while ensuring both attack effectiveness and stealthiness. Our PCDM incorporates an adjustable poisoning vector within the global context to precisely control the generation of poisoned data, with theoretical guarantees on attack performance. Furthermore, it employs a novel jumping diffusion strategy for lightweight and efficient poisoned data generation. We conduct the most systematic and broad experimental evaluation for FL poisoning attacks against various defenses, including advanced Byzantine robust aggregation mechanisms, on four open datasets: MNIST, Fashion-MNIST, CIFAR-10, CIFAR-100, and a real-world wireless-specific dataset VRAI. Our results demonstrate that PCDM is less likely to exhibit statistical anomalies compared with the state-of-the-art methods while more effectively degrading global FL performance, which poses a significant risk to data security in FL.

cs.CR↗
arXiv2026

AxiomOcean: Forecasting the Three-Dimensional Structure of the Upper Ocean

Short-term ocean forecast skill depends strongly on the three-dimensional ocean structure of the upper ocean, which governs stratification, subsurface heat storage, and the response of the ocean to atmospheric forcing. However, AI ocean forecasting models often fail to preserve this vertical structure, resulting in over-smoothed subsurface features and weak physical consistency under strong forcing. Here, we present AxiomOcean, a global AI ocean forecasting model that explicitly represents vertical hierarchy and cross-layer dependence within the water column. By combining a fully three-dimensional encoder-backbone-decoder architecture with surface atmospheric forcing, AxiomOcean jointly predicts upper-ocean temperature, salinity, and three-dimensional currents at global 1/12° resolution down to 643 m depth. In 10-day forecasts, AxiomOcean outperforms an advanced AI comparison model across variables and lead times, reducing day-1 RMSE by approximately 20 to 35% while maintaining higher anomaly correlation. The gain is not achieved through excessive smoothing: AxiomOcean better preserves eddy kinetic energy, temperature and salinity variance. Its advantage also extends through the water column and remains evident across the equatorial Pacific, Kuroshio Extension, and Southern Ocean, yielding a more realistic reconstruction of upper-ocean heat content. These results show that explicitly preserving upper-ocean three-dimensional structure can improve both forecast accuracy and physical fidelity in AI ocean prediction.

cs.LG↗
arXiv2026

Tandem: Riding Together with Large and Small Language Models for Efficient Reasoning

Recent advancements in large language models (LLMs) have catalyzed the rise of reasoning-intensive inference paradigms, where models perform explicit step-by-step reasoning before generating final answers. While such approaches improve answer quality and interpretability, they incur substantial computational overhead due to the prolonged generation sequences. In this paper, we propose Tandem, a novel collaborative framework that synergizes large and small language models (LLMs and SLMs) to achieve high-quality reasoning with significantly reduced computational cost. Specifically, the LLM serves as a strategic coordinator, efficiently generating a compact set of critical reasoning insights. These insights are then used to guide a smaller, more efficient SLM in executing the full reasoning process and delivering the final response. To balance efficiency and reliability, Tandem introduces a cost-aware termination mechanism that adaptively determines when sufficient reasoning guidance has been accumulated, enabling early stopping of the LLM's generation. Experiments on mathematical reasoning and code generation benchmarks demonstrate that Tandem reduces computational costs by approximately 40% compared to standalone LLM reasoning, while achieving superior or competitive performance. Furthermore, the sufficiency classifier trained on one domain transfers effectively to others without retraining. The code is available at: https://github.com/Applied-Machine-Learning-Lab/ACL2026_Tandem.

cs.AI↗
arXiv2026

StructMem: Structured Memory for Long-Horizon Behavior in LLMs

Long-term conversational agents need memory systems that capture relationships between events, not merely isolated facts, to support temporal reasoning and multi-hop question answering. Current approaches face a fundamental trade-off: flat memory is efficient but fails to model relational structure, while graph-based memory enables structured reasoning at the cost of expensive and fragile construction. To address these issues, we propose \textbf{StructMem}, a structure-enriched hierarchical memory framework that preserves event-level bindings and induces cross-event connections. By temporally anchoring dual perspectives and performing periodic semantic consolidation, StructMem improves temporal reasoning and multi-hop performance on \texttt{LoCoMo}, while substantially reducing token usage, API calls, and runtime compared to prior memory systems, see https://github.com/zjunlp/LightMem .

cs.CL↗
arXiv2026

VoxMind: An End-to-End Agentic Spoken Dialogue System

Recent end-to-end spoken dialogue models enable natural interaction. However, as user demands become increasingly complex, models that rely solely on conversational abilities often struggle to cope. Incorporating agentic capabilities is therefore essential: by enabling tool use, these models can extend their knowledge boundaries and better solve real-world tasks. Yet, existing research has largely concentrated on core perception and generation, with comparatively limited exploration of such tool-augmented extensions. To bridge this gap, we present VoxMind, an integrated framework designed to equip end-to-end spoken dialogue models with comprehensive agentic abilities. Leveraging our curated 470-hour AgentChat dataset, we incorporate a "Think-before-Speak" mechanism, enabling the model to internalize structured reasoning as a critical prerequisite for planning and response generation. Furthermore, to mitigate latency bottlenecks caused by large-scale tool integration, we propose a Multi-Agent Dynamic Tool Management architecture. By asynchronously delegating retrieval tasks to an auxiliary agent aligned with the main model's reasoning trajectory, this system effectively decouples inference latency from toolset size. Experimental results confirm that VoxMind achieves significant improvements in agent performance: compared with strong baselines, the task completion rate increases from 34.88% to 74.57%, outperforming Gemini-2.5-Pro on spoken agent tasks while preserving general conversational quality. The source code and associated data are publicly available at https://github.com/MM-Speech/VoxMind.

cs.SD↗
arXiv2026

Asking like Socrates: Socrates helps VLMs understand remote sensing images

Recent multimodal reasoning models, inspired by DeepSeek-R1, have significantly advanced vision-language systems. However, in remote sensing (RS) tasks, we observe widespread pseudo reasoning: models narrate the process of reasoning rather than genuinely reason toward the correct answer based on visual evidence. We attribute this to the Glance Effect, where a single, coarse perception of large-scale RS imagery results in incomplete understanding and reasoning based on linguistic self-consistency instead of visual evidence. To address this, we propose RS-EoT (Remote Sensing Evidence-of-Thought), a language-driven, iterative visual evidence-seeking paradigm. To instill this paradigm, we propose SocraticAgent, a self-play multi-agent system that synthesizes reasoning traces via alternating cycles of reasoning and visual inspection. To enhance and generalize these patterns, we propose a two-stage progressive RL strategy: first, RL on fine-grained Grounding tasks to enhance RS-EoT capabilities, followed by RL on RS VQA to generalize to broader understanding scenarios. Experiments show RS-EoT achieves state-of-the-art performance on multiple RS VQA and grounding benchmarks. Analyses reveal clear iterative cycles of reasoning and evidence seeking, confirming RS-EoT mitigates the Glance Effect and enables genuine evidence-grounded reasoning. Our code, data, and models are available at https://geox-lab.github.io/Asking_like_Socrates

cs.CV↗
arXiv2026

Optimizing Antenna Coding for Pixel Antenna Empowered SISO-OFDM Systems

This work investigates antenna coding optimization to enhance the channel capacity of single-input single-output orthogonal frequency division multiplexing (SISO-OFDM) systems empowered by highly reconfigurable pixel antennas. We first introduce the model for pixel antenna empowered SISO-OFDM systems using a beamspace channel representation. We next formulate the problem to maximize the channel capacity through jointly optimizing antenna coding and the power allocation across subcarriers and solve it by Successive Exhaustive Boolean Optimization (SEBO) and water-filling (WF) algorithm. To reduce computational complexity, a codebook-based approach is also proposed for antenna coding optimization. Simulation results show that the channel capacity of SISO-OFDM system across all signal-to-noise-ratio (SNR) regions considered can be enhanced through leveraging pixel antennas as compared to using conventional antenna with fixed configuration. This result demonstrates the effectiveness of antenna coding technology empowered by pixel antenna in enhancing SISO-OFDM systems.

eess.SP↗
arXiv2026

An Electricity Market with Reactive Power Trading: Incorporating Dynamic Operating Envelopes

Electricity market design that accounts for grid constraints such as voltage and thermal limits at the distribution level can increase opportunities for the grid integration of Distributed Energy Resources (DERs). In this paper, we consider rooftop solar backed by battery storage connected to a distribution grid. We design an electricity market to support customers sharing rooftop generation in excess of their energy demand, where customers earn a profit through peer-to-peer (P2P) energy trading. Our proposed electricity market also incorporates P2P reactive power trading to improve the voltage profile across a distribution feeder. We formulate the electricity market as an optimization-based problem, where voltage and thermal limits across a feeder are managed through the assignment of customer-specific dynamic operating envelopes (DOEs). The electricity market equilibrium is referred to as a competitive equilibrium, which is equivalent to a Nash equilibrium in a standard game. Our proposed market design is benchmarked using the IEEE 13-node test feeder.

eess.SY↗