arXiv subjects

Jing Li

Publications and source records attributed to Jing Li.

At least 19 recordsLinked to original sources

arXiv2026

MAGIC: Multi-task Gaussian process for joint imputation and classification in healthcare time series

Time series analysis has emerged as an important tool for improving patient diagnosis and management in healthcare applications. However, these applications commonly face two critical challenges: time misalignment and data sparsity. Traditional approaches address these issues through a two-step process of imputation followed by prediction. We propose MAGIC (Multi-tAsk Gaussian Process for Imputation and Classification), a novel unified framework that simultaneously performs class-informed missing value imputation and label prediction within a hierarchical multi-task Gaussian process coupled with functional logistic regression. To handle intractable likelihood components, MAGIC employs Taylor expansion approximations with bounded error analysis, and parameter estimation is performed using EM algorithm with block coordinate optimization supported by convergence analysis. We validate MAGIC through two healthcare applications: prediction of post-traumatic headache improvement following mild traumatic brain injury and prediction of in-hospital mortality within 48 hours after ICU admission. In both applications, MAGIC achieves superior predictive accuracy compared to existing methods. The ability to generate real-time and accurate predictions with limited samples facilitates early clinical assessment and treatment planning, enabling healthcare providers to make more informed treatment decisions.

stat.ML↗

arXiv2026

RubricAtlas: Mining Type-Level Rubrics from Expert Reports to Evaluate Deep Research Agents

Deep research agents conduct multi-step Web searches and synthesise retrieved information into analytical reports. Evaluating these outputs requires checking which analyses they deliver, beyond the presence of relevant facts. Query-specific binary rubrics make individual requirements explicit, but task-level checklists may leave the overall analytical structure underexamined. We introduce RubricAtlas, which derives type-level analytical requirements from expert reports on different objects within the same research type and organises detailed checks around recurring analytical questions. The requirements give related tasks a shared standard for what to analyse, without prescribing their answers. Our framework links binary checkpoints to report passages and evaluates completeness separately from argument validity. Built from 3,079 Chinese financial research reports, RubricAtlas covers 99 research types. Reference-report selection experiments show that individual reports cover only about half of their type's analytical questions, and that selecting requirements from different reports changes evaluation scores and agent rankings. Across ten agents, two judge pipelines achieve checkpoint agreement of $κ$ = 0.71 and agent-ranking agreement of Kendall $τ$ = 0.87. Controlled deletions reduce scores on the affected analyses, while padding produces no detectable increase. These findings support type-level rubrics for comparing the analytical completeness of Web deep research outputs, with individual decisions traceable to report evidence.

cs.AI↗

arXiv2026

AutoVend: Benchmarking LLM Agents on Long-Horizon Autonomous Vending Operations

Large language model (LLM) agents are extending agent-mediated commerce on the Web from shopping for consumers to operating businesses for sellers. In online-to-offline (O2O) commerce, decisions are issued through Web-based platforms but realized through costly physical execution. Evaluating such agents requires going beyond individual transactions to ask whether decisions made online turn into sustained economic returns. We introduce AutoVend, a benchmark in which one LLM agent manages 20 vending machines across five site types through a Web-based operating interface for 180 simulated days. Its data-grounded simulator combines a shared warehouse, procurement lead times, paid replenishment, inventory expiration, and recurring operating costs. We evaluate five LLM agents against a rule-based inventory-control baseline and diagnose their site coordination, execution, and performance over time. Four of five evaluated LLM agents fall short of this baseline, and returns per dollar of inference cost rank models differently from returns alone. Operational analyses show that gains in individual metrics do not consistently yield stronger economic growth, as inventory and execution costs erode these gains. Trajectory analysis further reveals difficulties in grounding feasible actions, managing inventory commitments, and converting feedback into effective policy revisions.

cs.AI↗

arXiv2026

LiveSearch: Questions as Executable Verifier Code for a Live Benchmark of Professional-Source Financial Search Agents

In finance, the answer that matters is usually the current one: prices, rates, and disclosures hold only as of a given day, so retrieving up-to-date facts is a core capability of a financial agent. Yet most financial benchmarks overlook timeliness. They fix each answer at annotation time, so the gold goes stale as markets move and, once leaked into training data, rewards recall rather than search and analysis. We introduce LiveSearch, a live benchmark that stores each question as executable verifier code: a typed program of calls to a professional financial data service, followed by a declarative extractor. The question is frozen, while its gold answer is determined by executing the program against professional financial data at evaluation time, and a daily job re-executes it with no human or language model in the loop. A memorized answer is thus likely outdated, and a model must reach the current one through tools and analysis. Construction is equally automatic: a language model only verbalizes programs enumerated from real tool responses and never chooses an answer. Under general web search, six contemporary tool-calling models reach 70.9% to 76.7% answer accuracy. Over three consecutive days, 70% of served questions change their gold value, and even under tolerance-aware matching, a literally reused earlier snapshot fails on a majority of questions within a single day. Per-question tracking uncovers dynamic shifts in correctness masked by aggregate accuracy.

cs.AI↗

arXiv2026

Wavefunction-templated metamaterials

Metamaterials, which store, guide and emit waves through their eigenmodes, have enabled the development of devices with unprecedented functionalities. Yet the spatial profiles of conventional metamaterials' eigenmodes are typically determined by their pre-determined structures (Hamiltonians). In this work, we introduce `wavefunction-templated metamaterials', a class of physically realizable metamaterials hosting diverse prescribed, spectrally isolated eigenstates. The resulting Hamiltonian geometries obey a boundary law, with the required nonlocality determined solely by single-contact boundary sites, independent of the state interior or lattice size. The sparse Hamiltonians are then Hermitianized, followed by spectral reshaping to open a protective gap around the target eigenstate. We validate our approach in discrete lattices and acoustic metamaterials. Although demonstrated in acoustics, the principle can be applied to thermal, electrical-circuit and photonic metamaterials.

physics.optics↗

arXiv2026

Not Every Agent Needs the Strongest Model: Unified Prompt-Model Optimization for Heterogeneous LLM-based Multi-Agent Systems

Large language model (LLM)-based multi-agent systems (MAS) rely on role-specific prompts and execution models to solve complex tasks collaboratively. However, prompt optimization typically keeps model assignments fixed, leaving differences in role requirements and their associated inference costs outside the search. We present UniMASO, a Unified framework for heterogeneous Multi-Agent System Optimization through the joint adaptation of prompts and model assignments. UniMASO evaluates candidate updates through local, downstream, and global feedback, using cost-aware Pareto search to retain useful quality-cost trade-offs. Topology-aware co-evolution then coordinates these updates according to agent roles and dependencies. Cross-level trajectory memory further supplies historical successes and failures to guide subsequent optimization. Experiments on five benchmarks and five execution topologies show that UniMASO achieves competitive task performance at substantially lower deployment cost than the strongest homogeneous configurations. Under centralized execution, it improves the average score from 69.79 to 70.40 while reducing token-weighted deployment cost by 85.2% relative to DeepSeek-V4-Pro with MASPO. These results support joint prompt-model adaptation as an effective approach to efficient multi-agent collaboration. The code is available at https://github.com/Achilles008/UniMASO_main.

cs.MA↗

arXiv2026

Diagnosing Personalization Failures in Mobile GUI Agents

Unlike conventional GUI agents, personalized GUI agents require user preferences to guide sequential execution over changing GUI states. This raises two challenges: determining HOW each preference should alter execution strategy and WHEN it should take effect along the interaction trajectory. Existing benchmarks primarily evaluate overall task success, offering limited insight into these questions. We introduce PrefDiag-Bench with two components: (1) a categorization of preferences into five types, characterizing HOW they affect execution; and (2) a stage-wise diagnostic framework with 1,664 probes, tracing WHEN preferences should take effect and where failures occur. We evaluate eleven GUI agent systems, including end-to-end agents and agentic workflows, under two forms of user context: explicit user profiles and behavioral memory logs. Although existing agents achieve relatively high completion rates on non-personalized tasks, performance is substantially lower on preference-personalized tasks. Even the best-performing model, Seed-2.0-Pro, achieves only 47.17% on Process preferences under explicit user profiles. Agents struggle particularly with Ranking and Process preferences. Further analysis reveals two key capability bottlenecks: selecting relevant evidence from user history and recognizing when preferences should take effect during execution. Controlled interventions that explicitly provide relevant evidence improve end-to-end task success by up to 7.06%. PrefDiag-Bench offers a more systematic evaluation of GUI agents' personalization capabilities beyond overall task success. Code and data are available at https://github.com/Zhixin-L/PrefDiag-Bench.

cs.AI↗

arXiv2026

SecLog-Bench: A Scalable Benchmark for Cybersecurity Log Reasoning with Scaffolded Evidence

Large language models (LLMs) have shown strong potential for cybersecurity, leading to growing interest in LLM-based automated cybersecurity log analysis. However, current datasets (1) are limited in scale; (2) rely mainly on single-source evidence; and (3) have only binary or multi-class attack labels, without dense annotations of the reasoning process. To address these limitations, SecLog-Bench has three key features. First, we build an extensible simulation platform that covers 29 attack scenarios, supports repeated execution and the integration of new scenarios, and enables scalable generation of cybersecurity logs with attack-stage labels. Second, we use multi-source logs, including audit logs, application logs, and network traffic, to jointly analyze cyber attacks. Third, we annotate each sample with the essential context for cybersecurity log reasoning and a structured step-by-step evidence chain. We evaluate eight mainstream LLMs on this benchmark for post-detection attack-progress reasoning. Results show that, when provided with scaffolded evidence, current LLMs exhibit modest capability in attack-progress reasoning, with the best model achieving only 44.44% accuracy and 45.07% Macro-F1. Their reasoning remains unreliable; i.e., the highest Stage Reasoning F1 (SRF1) score is only 54.87%. Moreover, even correct predictions may be accompanied by flawed reasoning, and reasoning quality is often misaligned with prediction accuracy.

cs.CR↗

arXiv2026

The SDSS-V Local Volume Mapper (LVM): Technical Overview

The Sloan Digital Sky Survey V (SDSS-V) Local Volume Mapper (LVM) is an ultra-wide integral field spectroscopic survey of the Milky Way, the Magellanic Clouds, and a sample of galaxies in the local volume. The LVM spatially resolved spectroscopy covers the bulk of the optical range (3600-9800 A) at medium spectral-resolution (R = 2200-5500), and maps the disks of these galaxies at high physical spatial resolution (< 1 pc in the MW and ~10 pc in the Magellanic Clouds). Observations are carried out with the LVM Instrument (LVM-I), a custom-built robotic telescope, instrument, and facility located at Las Campanas Observatory (LCO) in Chile. In this article we present a technical overview of the LVM project. We describe the hardware, the survey observing strategy, and the details of how science and calibration observations are carried out. We provide a high-level overview of the survey data flow, including the reduction and calibration of the LVM spectra, and the production of the main data products that are released by the project. We also present an assessment of the LVM data quality and instrumental stability based on science observations. The main goal of this article is to serve as a primer for LVM users, providing the technical information required for the adequate analysis, modeling, and interpretation of the LVM data.

astro-ph.IM↗

arXiv2026

All-Dielectric Tight-Binding Photonics via a Band-Gap Vacuum

Tight-binding photonics bridges condensed-matter physics and optics, providing a powerful framework for engineering band structures and topological phases with light. For photons, however, the vacuum is gapless, so resonators in air carry radiative tails that produce spurious long-range couplings. Here we embed resonators in a photonic band-gap vacuum (BgV), a photonic crystal whose band gap plays for photons the role vacuum plays for bound electrons. The BgV localizes each resonance exponentially, so designed short-range couplings define the lattice Hamiltonian. We thereby experimentally realize a faithful Benalcazar-Bernevig-Hughes Hamiltonian in an all-dielectric photonic crystal, resolving corner resonances and their orbitals in microwave measurements. Unlike the vacuum surrounding electrons, the BgV is itself designable, controlling relative couplings and closed-loop interference through its band dispersion and Bloch modes. Full-wave calculations and extracted models show that uniformly modifying the BgV, with all resonators fixed, switches the quantized polarization of a two-dimensional Su-Schrieffer-Heeger lattice and converts a zero-flux lattice into a $π$-flux one. We also numerically realize three-dimensional pyrochlore and hyperhoneycomb lattices with space groups $R\bar{3}m$ and $C2/c$, the latter a nodal-loop semimetal with topological drumhead surface states. These results establish the BgV as a designable vacuum for all-dielectric tight-binding photonics in two and three dimensions.

physics.optics↗

arXiv2026

Adaptive Dual-Teacher Distillation with Subnetwork Rectification for Bridging Semantic Gaps in Black-Box Domain Adaptation

Assuming that neither source data nor source model parameters are accessible, black-box domain adaptation (BBDA) represents a highly practical yet challenging setting, where transferable knowledge is limited to the predictions of a black-box source model. Existing approaches exploit such knowledge via pseudo-label refinement or by leveraging vision-language models (ViLs), but they often fail to reconcile the inherent discrepancy between task-specific knowledge from black-box models and language-aligned semantic priors of ViLs, resulting in suboptimal integration and degraded adaptation performance. To address this challenge, we propose adaptive Dual-Teacher Distillation with Subnetwork Rectification (DDSR), a framework that explicitly reconciles these complementary yet inconsistent knowledge sources. DDSR employs an adaptive prediction fusion strategy to integrate predictions from the black-box source model and a ViL, generating reliable pseudo-labels for the target domain. A subnetwork-based regularization mechanism mitigates overfitting to noisy supervision by enforcing output consistency and gradient divergency. Furthermore, progressively improved target predictions iteratively refine both pseudo-labels and ViL prompts, enhancing semantic alignment. Finally, class-wise prototypes are used to further optimize target predictions via self-training. Extensive experiments on multiple benchmark datasets demonstrate that DDSR consistently outperforms state-of-the-art methods, including those with access to source data or source model parameters.

cs.CV↗

arXiv2026

A plug-in estimator of the non-iterative conditional expectation g-formula using long short-term memory models

The non-iterative conditional expectation (NICE) plug-in estimator of the g-formula can be used to estimate the causal effects of sustained treatment strategies, even in the presence of treatment-confounder feedback, under the assumptions of exchangeability, positivity, and consistency. In practical applications, parametric models are commonly used to estimate the joint distribution of confounders, treatments, and outcomes over time. As an alternative to parametric models, long short-term memory (LSTM) models, designed to capture both short-term patterns and long-range dependencies in sequential data, may better accommodate complex temporal dependencies among time-varying confounders, treatments, and outcomes. However, the validity of statistical inferences of a plug-in estimator based on LSTM models has not yet been investigated. Here, we provide proof of principle that, under a complex data generation mechanism that resembles data from a cohort of persons with HIV and sufficiently large sample sizes, an LSTM-based NICE g-formula plug-in estimator results in less bias than parametric estimators while achieving near-nominal coverage for 95\% confidence intervals. This work is intended to spur the development of a theoretical framework that encompasses deep learning plug-in estimators for causal inference based on high-dimensional longitudinal data.

stat.ME↗

arXiv2026

Safety Alignment via Constrained Knowledge Unlearning

Despite significant progress in safety alignment, large language models (LLMs) remain susceptible to jailbreak attacks. Existing defense mechanisms have not fully deleted harmful knowledge in LLMs, which allows such attacks to bypass safeguards and produce harmful outputs. To address this challenge, we propose a novel safety alignment strategy, Constrained Knowledge Unlearning (CKU), which focuses on two primary objectives: knowledge localization and retention, and unlearning harmful knowledge. CKU works by scoring neurons in specific multilayer perceptron (MLP) layers to identify a subset U of neurons associated with useful knowledge. During the unlearning process, CKU prunes the gradients of neurons in U to preserve valuable knowledge while effectively mitigating harmful content. Experimental results demonstrate that CKU significantly enhances model safety without compromising overall performance, offering a superior balance between safety and utility compared to existing methods. Additionally, our analysis of neuron knowledge sensitivity across various MLP layers provides valuable insights into the mechanics of safety alignment and model knowledge editing.

cs.CL↗

arXiv2026

All-in-One Memory: Process-Rewarded Reinforcement Learning for Memory Agents

Large language models increasingly rely on external memory to answer queries grounded in long-horizon histories such as multi-session dialogues and lengthy documents. However, existing memory methods often optimize retrieval and memory construction separately, without a shared reward space for evaluating search, writing, and stopping decisions. To address this, we propose All-in-One Memory (AiO-Mem), which organizes these operations as transitions over a shared evidence state and directly evaluates the semantic quality of each action in a unified reward space. AiO-Mem compiles heterogeneous histories into source-traceable multi-view events, providing a shared substrate for retrieval and memory writing. A unified memory policy actor then rolls out complete interaction trajectories over this shared evidence state. A separate trajectory-conditioned process reward model is trained through evidence-grounded supervised fine-tuning and trajectory-level DPO to assign process scores to the collected trajectories. These scores are aggregated into an action-balanced trajectory reward and used by Process-Aligned DAPO (PA-DAPO) to optimize the memory policy. Across long-context memory benchmarks, AiO-Mem consistently improves answer accuracy over both static memory systems and RL-trained memory agents.

cs.AI↗

arXiv2026

Identify, Classify, Act: Per-Instruction Controlled Compliance for LLMs over External Data

LLMs over external data face a per-instruction tension: a single document or tool return often interleaves legitimate directives with adversarial prompt injections, and the model must follow the former while rejecting the latter. Existing defenses suppress or trust all external instructions, collapsing the safety-utility frontier and forcing a choice between over-refusal and over-compliance. We argue the missing capability is \textbf{per-instruction credit attribution} and formalize it as External-Data Instruction Classification, an identify-classify-act decision over each embedded directive. We propose \textbf{EDIC}, a decomposed reward framework for per-instruction compliance. A scene-grounded pipeline injects mixed beneficial and harmful instructions with counterfactual filtering, while a generative judge is initialized by SFT and refined with GRPO to attribute compliance at the instruction level. The judge then produces a reward that separately encourages beneficial compliance and penalizes harmful compliance for RL training. Across two backbones and five benchmarks, EDIC improves both safety and utility over the strongest prior defense, lifting harmful-rejection by up to 7.7\% and Arena-Hard-Auto utility by 12.5\% on Qwen3-4B. The code repository is hosted at https://github.com/yingyingxia666/EDIC.

cs.CL↗

arXiv2026

ProV-Mem: Proactive Memory for Streaming Video Question Answering

Proactive streaming video question answering requires a model to continuously monitor an incoming stream and autonomously decide when to speak, what to say, and whether to emit a response. Existing streaming systems face a fundamental difficulty: retaining long visual histories is costly, whereas discarding them can lead to missed events and repeated responses. We propose \textbf{ProV-Mem}, a training-free proactive memory framework built around a frozen VLM backbone. ProV-Mem decomposes the proactive decision into three lightweight memory tiers: a temporal-trigger memory controls when to query the backbone and recovers from prolonged silence, an instructional memory guides what to say through a fixed prompting prior for concise and visually grounded generation, and an episodic memory determines whether to emit by filtering near-duplicate responses. Under an idealized conditional-sufficiency assumption, we derive a stagewise approximation bound relating the decomposed process to a full-state joint policy. We further introduce Cost-Aware Area Under the Curve (\textbf{CAUC}), which extends PAUC by explicitly discounting redundant, overly frequent, and unnecessarily long emissions. Experiments on the four scenarios of ProactiveVideoQA show that ProV-Mem achieves the best CAUC across all scenarios and the best PAUC on WEB, EGO, and TV, while maintaining low per-response inference latency.

cs.CV↗

arXiv2026

Boundedness and exponential stabilization in a degenerate migration-consumption system with logistic damping

We study the no-flux problem for a taxis-type migration-consumption system with logistic damping in a bounded planar domain, where the motility function $ϕ$ vanishes at zero and degenerates algebraically of order $α\ge1$. We prove uniform boundedness in $L^\infty(Ω)\times W^{1,\infty}(Ω)$ for every fixed damping coefficient $b>0$, without any smallness assumption on the initial data, thus improving the previous $L^2$-boundedness results. For $α>1$, we further prove that the solution converges exponentially to the spatially homogeneous state $(a/b,0)$, with an explicit admissible rate. The proof combines a weighted Gagliardo--Nirenberg type inequality adapted to the degeneracy with the dissipative structure of the signal equation. Therefore logistic damping drives the system toward homogeneity, despite the loss of uniform elliptic regularization.

math.AP↗

arXiv2026

Trace-to-Architecture: Capability Bottleneck Diagnosis for Harness Revision

Improving an LLM agent requires knowing what to change when it fails. Similar failures can require different fixes, and resolving one problem can expose another. We introduce Trace-to-Architecture (T2A), a framework that uses execution trajectories to determine what to change in an agent's harness and when those changes apply. A development agent diagnoses failures using execution evidence from successful and failed tasks, revises the corresponding harness components, and then checks whether target tasks improve, selected successful tasks remain successful, and the edits are actually triggered. New execution evidence guides subsequent diagnosis and revision. Across ALFWorld, ScienceWorld, and AppWorld, T2A achieves the highest aggregate score for all three evaluated task agents, exceeding their respective strongest baselines by 6.2--30.3 points. We also integrate T2A's diagnosis and validation procedures into Meta-Harness, improving the resulting harnesses' mean ScienceWorld score by 21.6 points from the same starting harness.

cs.AI↗