arXiv subjects

Jie Yang

Publications and source records attributed to Jie Yang.

At least 19 recordsLinked to original sources

arXiv2026

Reliable Bistatic Ranging With Unreliable Communication Payloads

This paper addresses a key challenge in bistatic integrated sensing and communications (ISAC): \textit{How to achieve reliable ranging using communication data signals that cannot be reliably recovered?} In modern communication systems, data payloads typically occupy far more time-frequency resources than pilots, offering substantially more observations to improve sensing accuracy. Exploiting them requires recovering the transmitted symbols to form sensing references. However, propagation and reflection losses can severely weaken target echoes. The sensing receiver may therefore be unable to recover data transmitted at a high rate selected for communication users. We first show that, at low signal-to-noise ratios (SNRs), using hard symbol decisions as sensing references can introduce a persistent bias in the estimated channel impulse response (CIR). This bias produces an error floor and spurious peaks in the power delay profile (PDP) that persist even after coherent accumulation over a large number of observations. To address this limitation, we develop a pilot-initialized expectation-maximization (EM) estimator that iteratively refines the CIR using posterior symbol moments. Under local regularity conditions and successful initialization, we establish that the EM method converges to the local maximum-likelihood (ML) solution for a sufficiently large number of OFDM data symbols. The converged estimator is asymptotically unbiased, where the CIR estimation mean-square error (MSE) at each delay bin decays as $O(1/M)$ by accumulating $M$ data symbols. Thus, estimation error can vanish through sufficient accumulation even while symbol recovery remains highly unreliable. Numerical results validate the proposed method, demonstrating substantially higher target detection probabilities than pilot-only and hard-decision-based bistatic ranging under the same low-SNR conditions.

eess.SP↗

arXiv2026

Low-Terahertz Channel Performance in Falling Wheat for Moisture Sensing

Terahertz (THz) waves are highly sensitive to water and therefore offer a promising noncontact approach to wheat moisture sensing, but existing THz studies have been limited to static samples. This article investigates moisture sensing based on the performance of a THz channel in falling wheat, as encountered at grain intake. The channel power loss is measured at six frequencies from 120 to 320 GHz for three moisture contents and interpreted with a physics-based model that couples the double-Debye permittivity of water, Landau-Lifshitz-Looyenga mixing, and Lorenz-Mie extinction. The model reproduces the measured loss. Above the extinction maximum, the dielectric moisture response peaks where the wavelength approaches the kernel diameter, and that a two-frequency loss ratio removes the loading dependence. A channel-level analysis of QPSK and 16-QAM links shows that a 1 dB moisture-induced loss change alters the bit error rate (BER) by a factor of 3 to 30. THz communication channels at grain intake could therefore sense moisture as a byproduct of data transmission.

physics.app-ph↗

arXiv2026

MMCOMET: A Large-Scale Multimodal Commonsense Knowledge Graph for Visual Reasoning

Commonsense knowledge is widely used to support multimodal reasoning and generation, yet existing resources are predominantly text-only and lack grounding in visual context. This limitation often leads to the retrieval of semantically related but visually misaligned knowledge, and can degrade downstream performance. We introduce MMCOMET, a large-scale multimodal commonsense knowledge graph that integrates physical, social, and eventive knowledge with visual grounding. MMCOMET extends ATOMIC2020 by associating each commonsense triple with representative images through a hybrid alignment pipeline, resulting in over 900K multimodal tuples spanning 19 relation types. We further establish a unified evaluation framework across diverse vision-language tasks, including visual commonsense reasoning, visual question answering, visual storytelling, image editing, and image captioning. Our analysis shows that multimodal grounding provides more contextually aligned and reliable knowledge compared to both text-only knowledge graphs and LLM-based knowledge generation, particularly in reasoning-intensive scenarios. At the same time, we observe that the effectiveness of external knowledge is task-dependent, and naive knowledge integration can introduce noise in perception-dominated settings. To mitigate this, we propose an adaptive knowledge router that dynamically selects optimal paths or bypass knowledge in harmful settings. MMCOMET provides a scalable multimodal commonsense resource and a benchmark for studying the role of grounded knowledge in multimodal systems.

cs.AI↗

arXiv2026

SelfImprove-Bench: Evaluating Generalization in Self-Improving Agents

Recursively self-improving (RSI) systems are increasingly viewed as a promising means of accelerating progress toward artificial general intelligence. They aim to autonomously improve their capabilities by refining their working pipelines and recursively improving the strategies that guide subsequent refinements, with gains generalizing to unseen scenarios.However, existing evaluations lack reliable metrics to distinguish capability gains from scenario fitting, sustainable and extensible protocols for unseen tasks and difficulties, and fine-grained attribution of improvement to evolution traces. To this end, we propose SelfImprove-Bench,an RSI evaluation framework that organizes tasks into three primary capabilities (i.e., Exploring, Planning, Reasoning) as scenarios for guiding model evolution. We instantiated the tasks on games for controllability and reproducibility, and pairs them with scalable generation and orchestration driven by task templates and difficulty coefficients. Beyond task-level scores, SelfImprove-Bench organizes evaluation into generalization ladder levels G0 to G4 that measure how far gains transfer from training conditions. Experiments reveal that the evolutionary gains of existing RSI methods are inconsistent across generalization levels.

cs.AI↗

arXiv2026

EvoMCP: Co-evolving MCP agents and training data

Supervised fine-tuning of Model Context Protocol (MCP) agents is typically conducted on a static data distribution, which fails to adapt to the student's evolving capabilities and leaves newly exposed weaknesses insufficiently addressed. To address these challenges, we propose \method, an agent--data mutual evolution framework that enables MCP agents to improve through a closed loop of task evolution, trajectory evolution, and supervised fine-tuning. Specifically, \method\ assesses the student's established capabilities and capability gaps from rollout feedback to identify learning targets. Task evolution then translates these targets into executable tasks by adapting the instructions and observable environment states of compatible verified seed tasks while keeping their verifiers aligned. To provide reliable and efficient demonstrations for these tasks, trajectory evolution uses reusable skills to recover failed teacher rollouts and refine successful ones. The resulting task--trajectory pairs are used to fine-tune the student, while accepted tasks seed subsequent evolution. To close the loop, we re-evaluate the updated student and use its feedback to guide the next round of data evolution. Experiments on Toolathlon, MCPMark, and MCP-Atlas demonstrate consistent improvements over the initial student, yielding absolute gains of 16.5\%, 10.2\%, and 12.6\%, respectively. Project page: https://github.com/xagao/EvoMCP.

cs.AI↗

arXiv2026

RoSAMDepth: Robust Self-supervised Depth Estimation Leveraging Segment Anything Model

Robust depth estimation aims to maintain high-quality depths across diverse conditions. However, most existing methods estimate depth without taking into account the object-level information. As a result, the predicted depth may easily deviate within objects and become blurred under adverse conditions. To overcome this weakness, we propose RoSAMDepth, a novel framework that can assist robust self-supervised depth estimation in leveraging rich and diverse object-level priors from the Segment Anything Model (SAM). We focus on incorporating object-level information across three key aspects: a segment-guided representation contrasting method that injects object-level awareness into the feature representation space; an adaptive regional outlier masking strategy combined with a regional Gaussian likelihood loss that enforces regional depth smoothness; and an object-level reliability estimation strategy that mitigates the influence of unreliable supervision. Extensive experiments across multiple datasets and diverse weather conditions demonstrate that our method produces sharper, more accurate depth predictions, consistently outperforming state-of-the-art methods. Our code is available at https://github.com/xagao/RoSAMDepth.

cs.CV↗

arXiv2026

AdaGround: Token-Normalized Adaptive Zooming with Verifier-in-the-Loop Refinement for GUI Grounding

GUI grounding on high-resolution screens fails mostly on small targets, and zooming in is a common training-free remedy. However, most existing zooming pipelines do not match the zoom to the target's size, accept the zoomed answer unconditionally and never revisit a wrong answer. To address these issues, we propose AdaGround, a training-free framework with three components: (1) Token-Normalized Adaptive Zooming, which sets the magnification, field of view and zoom depth of each sample from the model's own estimate of the target size; (2) Verifier-in-the-Loop Refinement, which uses one marked-click Yes/No verifier to keep or reject each zoom, choose between zoom chains, trigger re-grounding and, with an existence query, decide when to abstain; and (3) Compositional Structure Grounding, which composes spreadsheet cells from their headers. On Qwen3-VL models, AdaGround surpasses state-of-the-art training-free zooming methods and many trained GUI models, improving accuracy over the backbone by 7.4 to 20.6 points on OSWorld-G and ScreenSpot-Pro, and its gains carry over to held-out UI-Vision. Code is available at https://github.com/bcwang-sjtu/AdaGround

cs.CV↗

arXiv2026

GUI-JEPA: Candidate-Conditioned JEPA Latents as a Fast Action Policy for Mobile GUI Agents

Mobile GUI agents are typically fine-tuned 7B-scale vision-language models (VLMs) that generate each action as text. However, these VLMs decode every action token by token, and even agents that score candidate actions keep a large language model or VLM in the loop, making them slow and memory-hungry. To address these issues, we propose GUI-JEPA, to our knowledge the first decoder-free GUI agent policy built on a Joint-Embedding Predictive Architecture (JEPA), which selects actions in the latent space of a JEPA encoder pretrained on unlabelled GUI screenshots. Its action- and region-conditioned latent predictor turns the screen latent into one latent per candidate action, letting lightweight heads score all candidates against the instruction in a single batched pass, without decoding a token. We further show that the scorer relies on this latent pathway and that the predictor is trained for the decision itself: a randomly initialised predictor does just as well as one pretrained on screen dynamics. On AndroidControl, GUI-JEPA is 2.8-4.0x faster on the same GPU than a Qwen2.5-VL-7B agent we fine-tune with the same inputs, while reaching 86-92% of its in-distribution accuracy with 6% of its parameters. Code is available at https://github.com/bcwang-sjtu/GUI-JEPA

cs.AI↗

arXiv2026

A-optimal Designs under Generalized Linear Models

Designing efficient experiments under practical constraints is critical in both scientific research and industrial practice. Focusing on minimizing the average variance of the parameter estimates, A-optimal designs show advantages in screening factors and reducing prediction errors. Compared with other criteria, however, algorithms and software for generating A-optimal designs are scarce. In this paper, we characterize A-optimal designs under generalized linear models theoretically and develop efficient algorithms for identifying them. When a predetermined finite set of experimental settings is given, we derive analytic solutions or establish necessary and sufficient conditions for obtaining A-optimal approximate allocations. We show that a lift-one algorithm based on our formulae outperforms commonly used algorithms for finding A-optimal allocations. When continuous factors or design regions get involved, we develop a ForLion algorithm to find A-optimal designs with mixed factors. Numerical studies on real-data experiments show that our algorithms can find highly efficient designs with reduced numbers of distinct experimental settings, which may save both experimental time and cost significantly. Along with a rounding-off algorithm that converts approximate allocations to exact ones, we demonstrate that A-optimal designs may provide more accurate parameter estimates than commonly used alternatives.

stat.ME↗

arXiv2026

PAFS: Perturbation-Aware Firing Scaffolds for Robust Event-Based SNNs

Event-based spiking neural networks (SNNs) encode information through temporal firing dynamics, but the same stateful computation makes their predictions sensitive to event loss, timestamp jitter, and speed variation. These perturbations alter evidence accumulation, yet late-layer firing trajectories can retain error-diagnostic structure that is absent from the final logits alone. We introduce Perturbation-Aware Firing Scaffolds (PAFS), a labeled post-training calibration method that exploits this structure without updating the pretrained backbone. PAFS associates standardized late-layer leaky integrate-and-fire trajectories with the class-probability residuals of an Adapter anchor, then applies singular value decomposition to extract compact residual-aligned channel directions. At inference, firing trajectories from the same backbone pass are projected onto this fixed low-rank scaffold and combined with anchor statistics to predict a residual logit correction. Across four matched dataset-backbone settings spanning three event datasets and two spiking Transformer architectures, PAFS improves corrupted-average and worst-condition accuracy over the corresponding Adapter anchors, with maximum gains of 3.06 and 3.65 percentage points. In the DailyAction-DVS/Spikingformer setting, the added trainable parameters amount to approximately 0.34% of the frozen backbone.

cs.NE↗

arXiv2026

One Act, Many Masks: Effect-Marginal Policy Optimization for Long-Horizon Agents

Recent advances in multi-turn reinforcement learning (RL) have substantially improved the performance of LLM agents on complex, long-horizon interactive tasks. However, despite progress in fine-grained credit assignment, policy optimization still applies these learning signals through response-specific probability ratios. Different textual responses can induce the same executable effect, yet credit differences arising from subsequent outcomes remain tied to their individual textual realizations. We identify this mismatch as credit-update misalignment. Our analysis reveals that credit-text pairing directly shapes policy-update directions, even when the policy context and executable effect are held fixed. Motivated by this finding, we propose Effect-Marginal Policy Optimization (EMPO), which aligns policy optimization with execution-level equivalence through an effect-level surrogate that couples updates across alternative textual realizations. EMPO preserves the original credit signals, integrates with both trajectory-level and step-level credit estimators, and requires no additional critics or rollouts. Experiments on ALFWorld, WebShop, and Sokoban show that EMPO outperforms strong baselines in both task performance and learning efficiency.

cs.LG↗

arXiv2026

SkillNEO: Non-Elitist Evolutionary Optimization of Agent Skills

Agent skills turn procedural knowledge into reusable artifacts, allowing LLM agents to accumulate and refine expertise through experience. Building on this, skill evolution iteratively improves skills using execution feedback. However, improvement may require several interdependent edits, so low-reward versions can still be worth developing. Requiring immediate reward gains can discard useful partial repairs before their benefits emerge. Non-elitist evolution allows new versions to receive further edits despite scoring below earlier ones. For agent skills, this calls for evidence of what to preserve and repair next. We introduce SkillNEO (Non-Elitist Evolutionary Optimization for Agent Skills), which separately selects a version to use and versions to keep improving. Comparing how related versions perform the same task reveals working steps to preserve and remaining failures to fix. It can also revisit and revise earlier versions using later evidence. Across 86 SkillsBench tasks in eight domains, including software engineering and office work, SkillNEO raises GPT-5.4 Mini's average reward from 49.1% to 66.4%, within GPT-5.5 and GPT-5.6 Luna's original-skill range. Mini executes tasks, proposes revisions, and edits skills; final evaluation uses the same task instances. Gains span all eight domains. The same skills improve stronger models without further editing. On the 41 tasks with modified skills, reuse improves Luna by 10.5 percentage points and places GPT-5.5 10.6 points above its published original-skill reference. Execution records show these skills can benefit stronger models even when Mini still fails some task requirements. Search histories also show lower-scoring versions leading to skills ultimately selected for use. Together, these findings show that skill evolution can turn a weaker model's partial progress into reusable procedural knowledge for improving more capable agents.

cs.AI↗

arXiv2026

PAINT: A Framework for Multilingual Pain Representation from Medical Licensing Exams to Real-World Clinical Texts

Pain is one of the most common and clinically important symptoms, reflecting disease onset, progression, and treatment response. However, pain is difficult to accurately capture and interpret from medical text, as its expression varies across languages, documentation types, and patient demographics, and is often embedded in terminology, negation, temporality, and contextual descriptions. Existing pain-related studies remain largely monolingual or limited to narrow clinical settings. In this study, we introduce PAINT, a multilingual framework for studying pain representation across medical exams and real-world clinical texts. PAINT includes nine pain-focused dataset partitions spanning English, Spanish, Portuguese, and Japanese, with clinician annotations in all four languages. We curated a structured schema covering pain relevance, pain presence, anatomical location, pain descriptors, and demographic information. To capture multilingual pain representations, we leverage large language models (LLMs) and adopt an ensemble strategy to improve performance. Ten clinicians across four languages validate the labels: the ensemble reaches macro-F1 of 0.91-1.00 on pain relevance and pain presence and 92-100% accuracy on demographics, with inter-annotator $κ$ of 0.82-1.00. We release the WorldMedQA-Pain ensemble labels, the extraction pipeline, validation scripts, and clinician annotation sheets for publicly licensed corpora, providing a foundation for pain-centered clinical NLP and for studying pain documentation across languages and healthcare contexts.

cs.CL↗

arXiv2026

UAV SAR Imaging via Diffusion Posterior Sampling under Non-Ideal Motion in ISAC Systems

Unmanned aerial vehicle-enabled synthetic aperture radar (SAR) offers significant potential for providing communication services to ground users while sensing the surrounding environment in low-altitude integrated sensing and communication (ISAC) systems. However, its practical deployment is challenged by known non-ULM trajectories and unknown position errors, which violate conventional SAR assumptions and cause severe image defocusing. We address these non-ideal impacts by formulating the ISAC SAR imaging as an exact linear inverse problem to accurately model arbitrary trajectories with multiple-input multiple-output orthogonal frequency division multiplexing waveforms. Based on this formulation, we develop a diffusion posterior sampling (DPS)-based super-resolution SAR framework, where a pre-trained diffusion model provides a data-driven image prior and the exact physical model enforces measurement consistency during posterior inference. The inference process is further accelerated for practical implementation. For unknown position errors, we develop a joint iterative DPS and motion compensation (MoCo) optimization algorithm with discrete cosine transform reparameterization and coarse-to-fine spectral scheduling. Simulations demonstrate that our framework achieves high-accuracy reconstruction across diverse trajectories and remarkable super-resolution, maintaining greater than 21.3 dB peak signal-to-noise ratio (PSNR) even under half physical resolution truncation. Under unknown motion errors, the joint DPS and MoCo algorithm achieves 23.65 dB PSNR with a residual position-error estimation error of 0.19 cm.

eess.SP↗

arXiv2026

PEARL: Process-Experience Aware Reinforcement Learning for Multi-Turn Agents

Terminal rewards make multi-turn agents trainable without step-level annotations, but group-relative advantages provide little guidance about which intermediate decisions to adjust. Failed trajectories also contain reasonable actions, while unfavorable decision patterns can recur as training proceeds. We introduce PEARL (Process-Experience Aware Reinforcement Learning), which improves local credit assignment without an additional critic. Given two continuations from the same history with different outcomes, PEARL locates their first divergence and builds a feature representation of each branch's local decision. It clusters these decision features in recent failure and success memories to identify recurring local patterns. Using this pattern memory, PEARL selectively strengthens negative advantages at lower-reward decisions, while leaving positive advantages and the base policy optimizer unchanged. On AppWorld, PEARL improves over ARPO on Normal and Challenge tasks at both 4B and 8B scales. On ALFWorld, it achieves 97.9% mean success at 4B, 2.7 points above the strongest baseline, while remaining competitive at 8B. Ablations of shaping location, success-bank filtering, and memory window support the value of selective, recurring-pattern-based updates.

cs.AI↗

arXiv2026

CiteVQA: Benchmarking Evidence Attribution for Trustworthy Document Intelligence

Multimodal Large Language Models (MLLMs) have significantly advanced document understanding, yet current Doc-VQA evaluations score only the final answer and leave the supporting evidence unchecked. This answer-only approach masks a critical failure mode: a model can land on the correct answer while grounding it in the wrong passage---a critical risk in high-stakes domains like law, finance, and medicine, where every conclusion must be traceable to a specific source region. To address this, we introduce CiteVQA, a benchmark that requires models to return \textit{element-level} bounding-box citations alongside each answer, evaluating both jointly. CiteVQA comprises 1,897 questions across 711 PDFs spanning seven domains and two languages, averaging 40.6 pages per document. To ensure fidelity and scalability, the ground-truth citations are generated by an automated pipeline---which identifies crucial evidence via masking ablation and enforces multi-stage quality control. At the core of our evaluation is Strict Attributed Accuracy (SAA), which credits a prediction only when the answer and the cited region are both correct. Auditing 20 MLLMs reveals a pervasive Attribution Hallucination: models frequently produce the right answer while citing the wrong region. The strongest system (Gemini-3.1-Pro-Preview) achieves an SAA of only 76.0, and the strongest open-source MLLM reaches just 22.5. Ultimately, towards trustworthy document intelligence, CiteVQA exposes a reliability gap that answer-only evaluations overlook, providing the instrumentation needed to close it. Our repository is available at https://github.com/opendatalab/CiteVQA.

cs.CL↗

arXiv2026

SDPAD: A Fully Spike-Driven Pipeline for End-to-End Autonomous Driving

End-to-end autonomous driving demands trajectory planners that are both highly accurate and cheap enough for edge deployment. State-of-the-art artificial neural network (ANN) planners meet the accuracy requirement at the cost of heavy dense computation, while spiking neural networks (SNNs)---though promising orders-of-magnitude energy savings through sparse, event-driven arithmetic---still lag far behind in planning accuracy. We present \textbf{SDPAD}, a fully spike-driven end-to-end planning pipeline that closes this gap. SDPAD converts a pre-trained ANN perception stack into integer-spike form via quantized ANN2SNN conversion, lifts multi-view images into the bird's-eye-view (BEV) space with a spike-driven-max (SDM) depth distribution (Spike-3D-Lift), and plans through the Spike-QFormer, a spiking query transformer in which ego, agent, and map queries distilled from the BEV scene are fused by learnable waypoint queries via cross-attention, followed by deformable spike-cross-attention refinement. Every operation is gated by integer spikes and inference is a single feed-forward pass without temporal simulation loops. On the nuScenes open-loop benchmark, SDPAD achieves an average $L_2$ error of 0.40\,m and a collision rate of 0.12\%, on par with strong ANN planners while consuming 69.9\,mJ---less than 2\% of recent ANN baselines. In closed-loop evaluation on the NAVSIM navtest split, SDPAD reaches 86.3 PDMS, surpassing the previous SNN planner SAD by 4.3 points and matching mainstream ANN planners at a fraction of their energy. To our knowledge, SDPAD is the first fully spike-driven planner evaluated in end-to-end autonomous driving, demonstrating that SNNs can rival dense ANNs in complex driving tasks.

cs.RO↗

arXiv2026

Regular integral models of Shimura varieties and a conjecture of Pappas

We construct regular integral models for a class of Shimura varieties of symplectic and orthogonal type with maximal parahoric level at an odd prime $p$. These models are defined over the ring of integers of the reflex field and have special fiber with normal crossings and irreducible components of multiplicity one or two. Our construction uses explicit equivariant modifications of the corresponding canonical local models for symplectic and split even orthogonal similitude groups, with the minuscule cocharacters corresponding to maximal isotropic Grassmannians. These modifications are obtained by successively blowing up Schubert varieties in the special fiber. This proves the equivariant modification conjecture of Pappas at maximal parahoric level in the above cases.

math.NT↗