arXiv subjects

Wei Fan

Publications and source records attributed to Wei Fan.

At least 19 recordsLinked to original sources

arXiv2026

Single-crystal X-ray diffraction of BC8 carbon near terapascal pressures

Whether carbon forms crystalline structures beyond diamond at terapascal pressures remains unresolved. Predicted for nearly four decades, the denser BC8 phase has lacked direct crystallographic evidence. Here we report the first direct crystallographic identification of BC8 carbon, formed as a single crystal from molten carbon during laser-driven decaying-shock compression of [100]-oriented diamond following loading above 1.1 TPa. Using in situ orientation-resolved single-crystal X-ray diffraction, we observe discrete Bragg spots whose spacings and angular relationships are reproduced by a common orientation matrix within each indexed BC8 reflection group, providing constraints beyond powder peak matching. Our single-crystal formation strategy and direct diffraction evidence resolve the decades-long quest for a post-diamond carbon structure, demonstrating how dynamic loading pathways can make predicted high-pressure phases experimentally accessible.

cond-mat.mtrl-sci↗

arXiv2026

TimesJev: Thinking Fast on Time Series with System 1 Sufficiency

Large language models (LLMs) have recently shown strong potential for time series reasoning, while models that reason over time series are usually with long chains of thought. However, this process often consumes thousands of tokens, directly leading to substantial latency during inference. In contrast, System 1 models such as Jev make decisions instantly and can naturally be an alternative, but raw time series are not compatible with the embedding space of these models, thus hindering the reasoning performance, especially when input series contain thousands of numeric points. In this paper, we make a very early attempt to propose a time series System 1 model, namely TimesJev. Specifically, we first formalize System 1 sufficiency in time series reasoning, which motivates a compressed representation for time series with key frequency patterns preserved. To construct this representation, we derive a dimension reduction objective that combines (i) a sufficiency divergence that measures the core information lost during compression, and (ii) a realizability divergence that quantifies the incompatibility of the System 1 model to the optimal prediction given this representation. We optimize this objective by learning time series compression with a gating mechanism for sufficiency, and then aligning the selected representation with the embedding space for realizability. Finally, experiments on 13 datasets demonstrate that TimesJev outperforms reasoning models by over 62% with 62.9% less inference time than their fine-tuned versions on average.

cs.AI↗

arXiv2026

CoworkEnv: Synthesizing Training Environments for Long-Horizon Cowork Agents

AI coworkers must coordinate tools, communicate across turns, and act on changing workspace state while respecting workflow constraints. Constructing training experience for these capabilities requires coherent requests, service records, tool interfaces, and grading checks. We present CoworkEnv, a capability-guided framework for synthesizing executable training environments for AI coworkers. Workflow requirements guide task construction, and programmatic validation with feedback repair checks generated specifications and grading behavior. Teacher executions yield curated demonstrations for supervised fine-tuning (SFT), while student interactions provide rollouts and rewards for reinforcement learning (RL). Using these resources, we train compact PocketWorker models. For Qwen3-VL-8B-Instruct, SFT raises the mean score from 46.6 to 64.1 on Claw-Eval, including a 31.9-point gain on Multi-turn tasks. An SFT-only extension to ClawMark demonstrates applicability across coworker benchmarks. We further compare RL training dynamics across configurations that vary reward processing and credit assignment. Improving training rewards provide preliminary evidence that the constructed environments support policy optimization.

cs.AI↗

arXiv2026

Paloma: Phase-Conditioned Residual Modulation for Time Series Forecasting

Explicitly modeling periodicity has proven to be an efficient approach for time series forecasting, where the residual, obtained after removing the periodic component from the time series, is typically treated as unstructured noise. However, in this work, we show that such residuals still exhibit structured, \textit{phase-dependent} deviations, as supported by both empirical observations and theoretical analysis. This suggests that residuals should not be treated merely as unstructured noise, but instead as \textit{phase-conditioned signals} whose variations arise in both direction and magnitude. Motivated by this perspective and inspired by phase and amplitude modulation in classical signal processing, we propose \textit{Paloma}, a simple yet effective framework for modeling residual deviations. Specifically, Paloma comprises two complementary modules: \textit{a Phase Rotation module}, which performs phase-conditioned rotations to capture directional changes in the feature space, and \textit{an Amplitude Scaling module}, which applies phase-conditioned affine transformations to model magnitude variations. We further abstract the two modules into a general-purpose plugin, termed the \textit{Paloma technique}, which can be readily integrated into existing forecasting architectures to modulate input sequences in a phase-aware manner, enabling backbone models to capture phase-dependent variations in both direction and magnitude. Extensive experiments on twelve datasets demonstrate that Paloma achieves state-of-the-art and robust forecasting performance. Additional results show that the Paloma technique consistently improves a wide range of baseline models, validating its effectiveness as a general-purpose component. Our Code is available at \url{https://github.com/TSTS13/Paloma}.

cs.LG↗

arXiv2026

LoOPSD: On-Policy Self-Distillation with Looped Teachers

On-policy self-distillation improves a model using guidance from the model itself, but it still requires a teacher that provides more useful predictions than the student. Existing approaches obtain such guidance by giving the teacher additional context or instructions. We ask whether it can instead come from changing how the same model uses its own layers. To address this, we introduce LoOPSD, which searches over a model's layer execution order to construct a stronger teacher and distills its advantage into the student. Specifically, we keep the model weights unchanged while searching for layers to repeat, and use the resulting teacher's predictions on student-generated prefixes to guide learning. We then alternate search and distillation, rebuilding the teacher from the updated student in each round so that the student progressively absorbs the gains from alternative layer execution. Experiments on mathematical and logical reasoning tasks show improvements across models. On GSM8K with a 400-token thinking budget, LoOPSD raises Qwen3-1.7B accuracy from 42.30% to 79.61%; subsequent distillation rounds yield further gains beyond the first round. Together, these findings show that by searching layer execution to construct teachers, we can turn a model's additional computation into improved student capabilities, opening a new avenue for self-distillation to derive supervision from the model's own computation.

cs.LG↗

arXiv2026

Retarded Green Function For Bopp--Podolsky Electrodynamics in Four-Dimensional De Sitter Spacetime

We study classical Bopp--Podolsky (BP) electrodynamics on four-dimensional de Sitter spacetime, focusing on its transverse retarded Green function and gauge-invariant electromagnetic response. Because de Sitter spacetime is maximally symmetric, its vector Green functions admit a covariant expansion in a finite basis of invariant bitensors, with scalar coefficients depending only on the invariant separation. For conserved sources, the fourth-order field operator factorizes into Maxwell and massive-vector operators, yielding the BP retarded kernel as the difference of their transverse Green functions. Using the invariant bitensor decomposition, we reduce the propagator equations to a hypergeometric master equation and determine the distributional structure of its solution. The common leading short-distance singularity of the Maxwell and massive-vector sectors cancels, leaving a logarithmic master singularity and a softened vector-potential wavefront. After one curl, the gauge-invariant field-strength kernel contains a null-cone delta distribution and a regular timelike tail. The smooth Maxwell master tail is annihilated by the curl, so the interior electromagnetic tail is generated entirely by the massive sector. We also derive an exact horizon-regular static potential, demonstrate the cancellation of the Coulomb divergence, recover the flat-space BP potential, and formulate the generalized Gauss law. On global de Sitter spacetime, the compactness of the spatial sections requires the total charge to vanish; we implement this constraint using an antipodal charge configuration with smooth factorized initial data. The analysis is classical and does not assume a quantum Lee--Wick contour or a preferred de Sitter vacuum.

hep-th↗

arXiv2026

SkillHorizon: Constructing Long-Horizon Agent Tasks through Skill Composition

Long-horizon agents must complete workflows in which later activities depend on earlier outputs and changes to shared state. A key challenge for evaluation is to expand task requirements and their dependencies while keeping tasks executable and outcomes verifiable. We introduce SkillHorizon, a benchmark that constructs long-horizon tasks through skill composition in seven business environments. Its generators use domain-specific rules to turn reusable skill procedures into concrete business activities. These activities are connected through input-output dependencies, shared resources, and state updates, allowing workload and coordination demands to grow together. Successful reference executions validate task feasibility and measure interaction workload, while programmatic evaluators assess local completion and whole-task success. We evaluate five language models on 1,050 tasks with reference interaction workloads ranging from 64K to 1M tokens. Across these scales, mean local-goal completion falls from 90.4% to 46.8%, as completed work fails to keep pace with growing requirements. At larger workloads, agents more often submit unsuccessful final answers with substantial work unfinished and tool-call allowance remaining. Trace analysis also identifies cases where scripts execute successfully but their outputs fail downstream business checks. SkillHorizon provides a configurable testbed for studying sustained progress, dependency coordination, and complete task delivery in growing workflows. Our code is available at https://github.com/KnowComp/SkillHorizon.

cs.AI↗

arXiv2026

Electron- and hole-doped superconductivity in two-dimensional electride ZrX$_2$ (X = Cl, Br, I)

Two-dimensional (2D) superconducting systems have recently emerged as a frontier in condensed matter physics, driven by their unique physical properties and potential applications. Using first-principles calculations, we show that monolayer ZrX$_2$ (X = Cl, Br, I), which is a family of 2D semiconducting electrides, hosts phonon-mediated superconductivity under both electron and hole dopings. Electron doping in ZrX$_{2}$ can achieve high superconducting transition temperatures at very low concentrations, with ZrBr$_2$ reaching $T_c = 30.0$ K at 0.01 \textit{e}/f.u., followed by ZrCl$_2$ (22.5 K at 0.05 \textit{e}/f.u.) and ZrI$_2$ (18.1 K at 0.01 \textit{e}/f.u.), for which the high $T_c$ originates from the flat conduction band minimum, with Van Hove singularities nearby enhancing the density of states. In contrast, hole doping requires higher concentrations (0.2--0.25 \textit{h}/f.u.) to activate superconductivity, yielding $T_c \sim 2$ K, and approaches a magnetic instability at elevated doping levels. By comparing with InSe and LaBr$_2$, we identify the spatial location and orbital character of excess electrons as key factors determining the strength of hole-doped superconductivity and the nature of competing orders. These findings establish monolayer ZrX$_2$ as a platform for bipolar superconductivity in a 2D electride and provide insights into flat-band engineering for high-$T_c$ superconductors.

cond-mat.supr-con↗

arXiv2026

Understanding Errors in LLM-Based Question Answering over Imperfect Tables

Answering questions over imperfect tables requires handling errors that can affect the answer. We investigate two challenges for large language models (LLMs): whether error discovery depends on where errors appear in a table, and whether providing their locations is sufficient for accurate question answering (QA). Using human-reviewed instances from RADAR-T, we conduct controlled studies across three LLMs by varying row order and comparing original, error-marked, and repaired tables. First, reordering rows changes error discovery even when the table contents and gold answer remain unchanged. During direct inspection, LLMs are more likely to discover all rows containing relevant errors when these rows appear later in the table or are grouped more closely together. Second, providing verified error locations alone is insufficient for accurate QA: with code execution, accuracy on repaired tables exceeds that on error-marked tables by 39.0-59.1 percentage points across the three LLMs. As a practical application of these findings, we combine error discovery across shuffled table views with explicit guidance for verifying and handling the reported errors in a simple workflow, Geometry-Balanced Discovery and Intervention (GBDI). On RADAR-T, GBDI improves QA accuracy by 3.8-18.5 percentage points over a code-agent baseline across five LLMs (paired 95% confidence intervals exclude zero for four), at the cost of additional inference. These results highlight the importance of both reliable error discovery and effective error handling in QA over imperfect tables. Code is available at https://github.com/645-t/GBDI-ICLR-2027.

cs.CL↗

arXiv2026

Multi-feed Plane Wave Generator for Compact OTA Multi-Target Emulation With Joint Angle-Delay-Doppler Control

Over-the-air (OTA) radar target emulation is an important method for evaluating the sensing capabilities of integrated sensing and communication (ISAC) and radar devices. Conventional single-feed compact antenna test ranges (CATRs) are generally limited to a single target direction, while multi-feed angular-synthesis methods commonly used in radar testing still require far-field illumination over the device-under-test (DUT) aperture. To overcome these limitations, this paper proposes a radar target OTA emulation framework based on multi-feed plane wave generator (PWG). The key capability is the simultaneous emulation of spatially close targets from different directions with joint angle--range--Doppler control. Target range is emulated through the corresponding delay, while Doppler and the complex echo coefficient are controlled by a channel emulator. Meanwhile, a multi-input amplitude and phase matrix (APM) maps each independent target-signal path to a dedicated PWG excitation vector that synthesizes the corresponding incident plane wave over the same test zone. This enables multiple target directions to be reproduced as local plane waves at a test distance substantially shorter than the conventional far-field requirement. The framework is experimentally validated at 3~GHz using a real $1\times11$ phased-array DUT. The PWG--DUT separation is 1.6~m, substantially shorter than the 7.2~m Fraunhofer distance of the DUT. At this separation, the synthesized $0^{\circ}$ and $10^{\circ}$ incident waves each exhibit an amplitude peak-to-peak error of 1.0~dB. Across two four-target angle--delay cases and one two-target angle--velocity case, the estimated target parameters show angular deviations within $1^{\circ}$, interpolated delay errors within 2.4~ns, and velocity errors within 0.15~m/s.

eess.SP↗

arXiv2026

Realistic Characterization and Modeling of Vehicle-Mounted Antenna Radiation Patterns Based on Large-Scale Full-Vehicle Measurements

Vehicle-mounted antenna radiation patterns can be significantly altered after installation due to interactions with the vehicle body. However, existing studies mainly focus on simulations or isolated antenna characterization, while large-scale experimental investigations remain limited. This paper presents a comprehensive experimental analysis of vehicle-mounted antenna radiation patterns based on full-vehicle spherical near-field measurements. A large-scale dataset covering ten commercial vehicle platforms, multiple antenna mounting locations, and various wireless functions, including 4G, 5G, V2X, and GNSS, is established under a unified measurement framework. Based on the measured results, representative case studies and statistical characterization are performed to investigate the impacts of vehicle installation on antenna radiation behavior. The results reveal substantial variations across mounting locations and wireless functions, highlighting the strong influence of vehicle integration on realistic antenna performance. Furthermore, a novel statistical model is proposed to generate realistic vehicle-mounted antenna patterns using a small set of physically interpretable parameters. The obtained dataset, characterization results, and modeling approach provide a foundation for realistic antenna evaluation and wireless system simulations in vehicular communication applications.

eess.SP↗

arXiv2026

Distance-aware Self-adaptive Graph Convolution for Fine-grained Hierarchical Recommendation

Graph Convolutional Networks (GCNs) are widely used to improve recommendation accuracy and performance by effectively learning the representations of user and item nodes. However, two major challenges remain: (1) the lack of further optimization in the graph representation structure and (2) insufficient attention given to the varying contributions of different convolutional layers.This paper proposes SAGCN, a distance-based adaptive hierarchical aggregation method that refines the aggregation process through differentiated representation metrics. SAGCN introduces a detailed approach to multilayer information aggregation and representation space optimization, enabling the model to learn hierarchical embedding weights based on the distance between hierarchical representations. This innovation allows for more precise cross-layer information aggregation, improves the model's ability to capture hierarchical embeddings, and optimizes the representation space structure. Additionally, the objective loss function is refined to better align with recommendation tasks.Extensive experiments conducted on four real-world datasets demonstrate significant improvements, including over a 5% increase on Yelp and a 5.58% increase in Recall@10 on the ML_1M dataset.

cs.IR↗

arXiv2026

ClinLens: Towards Long-Horizon LLM Agents for Longitudinal Multimodal Clinical Data Science

Automating clinical data science requires agents to translate patient-centric objectives into executable workflows grounded in longitudinal multimodal evidence. Existing benchmarks provide limited coverage of two essential capabilities: grounding analyses in the correct patient episode and harmonizing multimodal observations. We introduce ClinLens, a patient-centric benchmark for evaluating large language model (LLM) agents on long-horizon clinical data science workflows. ClinLens comprises 126 executable tasks spanning five data modalities, four analytical scopes, and five analytical capabilities. The tasks assess agents' ability to maintain a coherent patient-specific analytical context across interdependent stages of evidence grounding, multimodal integration, and analytical pipeline generation and execution. We evaluate general-purpose, coding, and biomedical agents, with task runs averaging 36.93 steps and 8.06 minutes. The strongest evaluated agent achieves an execution success rate of 86.4% and a final-answer accuracy of 54.0%, but a strict pass rate of only 29.4%. Failure analysis identifies patient-episode grounding as the most frequently recorded failure category under the evaluation procedure. By jointly evaluating analytical outcomes and the validity of the supporting clinical context, ClinLens provides a testbed for advancing reliable agents for longitudinal clinical data science.

cs.AI↗

arXiv2026

Verifiable Hidden Dynamics Play: Generating Agentic RL Environments from Solved Mechanisms

Language-model agents increasingly face long-horizon tasks with evolving state, interdependent decisions, and delayed outcomes. Scaling their training requires diverse agentic environments, dependable outcome signals, and low extension cost. Existing generation pipelines commonly construct an environment before defining its outcome rule or annotating its trajectories, leaving dynamics and evaluation to be aligned post hoc. VHD-Play reverses this dependency by sampling and solving a mathematical model before a corpus-grounded setter renders its decision process as stateful tools. The executable dynamics and trajectory-scoring reference are inherited from the same solved model. The pipeline produces 3,300 diverse agentic environments at a cost of a few cents each. Training Qwen3.6-35B-A3B on three families raises its mean agentic score from 0.204 to 0.815 in a five-family diagnostic. Gains also appear on held-out instances from all three training families and eight unseen mechanism families, then extend beyond the generated substrate to external benchmarks for general function calling, travel planning, and 365-day e-commerce. On E-Commerce Bench, the trained checkpoint completes every run without bankruptcy and exceeds Qwen3.7-Max. We compare written-out problems with stateful versions that reveal or hide their parameters. The comparison shows that most of the learnable gap lies in stateful interaction rather than underlying problem solving. A frozen 35B setter realizes larger environments, and scale-matched training retains gains as mechanism size and horizon grow, indicating the potential for an evolving training substrate.

cs.AI↗

arXiv2026

Regime transitions of dynamos driven by fingering double-diffusive convection

Long-term cooling of planetary interiors is expected to progressively establish thermally stable stratification in the liquid outer core, yet its influence on dynamo action remains incompletely understood. Here we present three-dimensional numerical simulations of fingering double-diffusive convection-driven dynamos to investigate how progressively strengthening thermally stable stratification, modifies compositional convection driven flows and dynamos. Our results show that dynamo evolution is governed by a competition between Lorentz-force regulation and stratification-induced flow reorganisation. When stratification is weak, the system remains in a strong-field regime characterised by a relatively stable, dipole-dominated magnetic field, despite reduced flow intensity. As thermally stable stratification strengthens, systematic changes in flow morphology emerge, including the development of prograde equatorial zonal flows. Once a critical stratification strength is exceeded, flow reorganisation becomes dominant, leading to an abrupt transition to a weak-field dynamo accompanied by a sharp decline in magnetic energy. In the weak-field regime, magnetic field intensity exhibits enhanced temporal variability and continues to decay as stratification further strengthens, eventually resulting in dynamo cessation even when compositional buoyancy persists. These results indicate that thermally stable stratification may play an important role in regulating planetary dynamos and provide new insight into the long-term magnetic evolution of terrestrial planets.

astro-ph.EP↗

arXiv2026

Can LLMs Time Travel? Enhancing Temporal Consistency in Legal Agentic Search through Reinforcement Learning

While large language models (LLMs) augmented with agentic search capabilities show promise for legal reasoning, they overlook a fundamental constraint that applicable law must match the temporal context of each case, as retroactive application of statutes violates core legal principles and leads to erroneous conclusions. Our observations reveal that current legal LLMs suffer from temporal bias anchored to their training cutoff, while search agents rarely incorporate temporal constraints into queries, and that web search alone cannot provide the precise statute and precedent citations that legal reasoning demands. To address these challenges, we propose LegalSearch-R1, an end-to-end reinforcement learning framework that pairs local statute RAG for precise article matching with online web search for broader legal knowledge, trained on temporally-indexed data spanning multiple amendment periods to enforce temporal consistency. Extensive experiments on our benchmark covering 13 legal tasks demonstrate that our 7B-parameter agent outperforms state-of-the-art deep research frameworks and specialized legal LLMs by 12.9% to 29.8%, surpasses baselines by 57.7% to 80.3% on temporal consistency, and exhibits robust out-of-domain generalization. The code and data are available at https://github.com/AlexFanw/LegalSearch-R1.

cs.CL↗

arXiv2026

Emulation of Closely Spaced Spatial Sensing Targets Using a Parabolic Cylindrical Compact Range with a Multi-Feed Amplitude-Phase Controlled Linear Array

Integrated sensing and communication (ISAC) base stations (BSs) are emerging as critical infrastructure for 6G networks, requiring advanced over-the-air (OTA) testing methodologies that can emulate dynamic, multi-target scenarios under far field (FF) conditions. However, existing testing solutions often suffer from limited angular flexibility, insufficient bandwidth, and prohibitive costs. {To address these challenges, this paper proposes an OTA testing system that integrates a radar target simulator (RTS), a multi-port amplitude and phase modulation (APM) network, and a multi-feed linear array with a single parabolic cylindrical compact range (SPCCR).} The proposed system is capable of emulating closely spaced spatial targets with varying incident azimuth angles. This setup meets the essential conditions for ISAC BS testing, including FF characteristics, wideband and multi-band compatibility, and the capability for sensing multiple dynamic closed-spaced targets. The feasibility of the proposed method is validated through both FEKO simulation and measurement. Experimental results demonstrate that the system successfully emulates multiple closely spaced targets within an azimuth range of $\pm 10^\circ$ with high emulation accuracy while maintaining excellent quiet zone (QZ) performance.

eess.SP↗

arXiv2026

Hierarchical and Permutation-Invariant Feature Transformation Learning via Policy-Guided Embedding Search

Feature transformation improves predictive performance on tabular data by constructing informative abstractions from raw features. Recent generative approaches encode transformation knowledge into continuous embedding spaces for efficient exploration of candidate strategies, but face three key limitations: (1) overlooking hierarchical relationships between low-level features, operations, and high-level abstractions; (2) enforcing order-sensitive embeddings on inherently permutation-invariant transformation sequences, thereby introducing systematic bias; and (3) relying on gradient-based search, which is ill-suited to non-convex transformation spaces. We propose a framework with two complementary components. First, a permutation-invariant hierarchical module captures interactions across features, operations, and abstraction levels, with a self-attention pooling mechanism that maps semantically equivalent structures to consistent embeddings aligned with downstream performance. Second, a policy-guided multi-objective reinforcement learning strategy initializes the search from empirically strong seeds and jointly optimizes predictive accuracy and transformation efficiency. Extensive experiments on diverse tabular benchmarks demonstrate the effectiveness and robustness of our framework against strong baselines. Our code and data are publicly available at: https://github.com/RayLiu1103/PHER.

cs.LG↗