Showing 1–50 of 494 results for author: Wen, Q

An Interactive Agent for Requirement-Driven Candidate Sourcing

Authors: Yuanpeng He, Fangjing Li, Xiangyu Ru, Kexin Sun, Kun Yang, Lijian Li, Chi-Man Pun, Qingsong Wen, Wenpin Jiao, Mingkai Guo, Yirong Feng, Daiheng Gao, Zhi Jin

Abstract: Finding people from a natural-language description (``ML engineers transitioning to research roles in biotech'') is increasingly delegated to LLM agents and framed as information retrieval. We argue that it is fundamentally a requirements engineering task: such a request is an under-determined requirement with implicit constraints, many valid answers, and no acceptance criterion, so useful answers… ▽ More Finding people from a natural-language description (``ML engineers transitioning to research roles in biotech'') is increasingly delegated to LLM agents and framed as information retrieval. We argue that it is fundamentally a requirements engineering task: such a request is an under-determined requirement with implicit constraints, many valid answers, and no acceptance criterion, so useful answers require eliciting, validating, and verifying the requirement before search can matter. We present \sys{}, to our knowledge the first interactive, requirements-driven candidate-sourcing agent (it elicits, validates, retrieves, and verifies a vague people-request into a justified slate through bounded elicitation, workflow templates, a two-stage commit protocol, and bidirectional termination guards) and \bench{}, a benchmark that runs the requirements lifecycle (criteria-anchored validation, multi-model evidence-grounded oracle construction, and cost-aware verification). Across $21$ systems and all $691$ requirements, \sys{} dominates breadth ($100%$ coverage at $2.5\times$ the yield) and is \emph{near-orthogonal} to the field, with $90%$ of the people it returns are surfaced by \emph{none} of $20$ strong LLM-plus-web baselines combined. Beyond breadth, an evidence-grounded judging of every system shows \sys{} \emph{recalls} the most relevant real people: $0.241$ of the union pool, $1.9\times$ the next system, with a bootstrap $95%$ interval disjoint from every baseline. \sys{} is thus the strongest \emph{sourcing} engine (the deepest real, reachable candidate pool), while precision-ranking LLMs serve as~complementary verifiers. △ Less

Submitted 24 August, 2026; originally announced August 2026.

Comments: 12 pages

Mint-Agent: Introducing Finance-Native Agentic Foundation Models

Authors: Mint-Agent Team, Kun Wang, Gavin Zhang, Yaze Geng, Lei Tang, Yaoyang Yi, Zonghan Wu, Yifan Hu, Qingsong Wen, Yilei Shao

Abstract: Financial agents must do more than recall domain knowledge: they must be both reliable, executing precise operations over grounded evidence, and executive, sustaining long-horizon research whose conclusions remain auditable. We present Mint-Agent, a family of finance-native agentic models designed around these two scales of financial intelligence. Mint-Agent is built upon three pillars: data, harn… ▽ More Financial agents must do more than recall domain knowledge: they must be both reliable, executing precise operations over grounded evidence, and executive, sustaining long-horizon research whose conclusions remain auditable. We present Mint-Agent, a family of finance-native agentic models designed around these two scales of financial intelligence. Mint-Agent is built upon three pillars: data, harness, and algorithm. Our data engine constructs clean, specialized tasks for atomic financial capabilities and long-horizon agentic execution from real-world financial sources. MintHarness enables stable interaction with open-ended environments and maintains auditable evidence trails across extended research trajectories. Our training recipe combines SFT, critical-step OPD, and RLVR to develop separate financial reasoning and agentic execution experts, which are then unified through model merging and multi-teacher on-policy distillation into compact, general-purpose financial agents. This pipeline yields two flagship models, Mint-Cu (9B) and Mint-Ag (27B). Across professional financial benchmarks, our models demonstrate two defining strengths: (1) Reliability: Mint-Ag achieves 98.33% on RFC-Bench, surpassing GPT-5.6-Sol and Claude-Opus-4.8 by 3.66 and 3.00 points; and (2) Executability: Mint-Cu reaches 69.86% on FinSearchComp T2, outperforming Agents-A1-35B and Nex-N2-mini by 22.83 and 12.78 points, while Mint-Ag achieves 76.00% and 60.49% on FinanceAgentBench v1.1 and v2, respectively. These results establish a path toward trustworthy financial intelligence in which domain expertise, long-horizon execution, and auditable evidence are jointly engineered as a unified foundation for frontier agentic models. △ Less

Submitted 21 August, 2026; v1 submitted 17 August, 2026; originally announced August 2026.

HalluTracer: Hallucination Detection via Depth-Averaging Truth Signals

Authors: Zhihao Guo, Zonghan Wu, Huan Huo, DaYong Ye, Junwei Zhang, Weiran Yao, Zhiwei Liu, Qingsong Wen, Yilei Shao

Abstract: Even well-aligned large language models confidently generate factually incorrect text, making hallucination a persistent reliability risk in high-stakes deployments. These models nonetheless carry linearly separable truthfulness signals in their internal representations. Existing white-box detectors, however, collapse this evidence to isolated components or a single depth, discarding discriminativ… ▽ More Even well-aligned large language models confidently generate factually incorrect text, making hallucination a persistent reliability risk in high-stakes deployments. These models nonetheless carry linearly separable truthfulness signals in their internal representations. Existing white-box detectors, however, collapse this evidence to isolated components or a single depth, discarding discriminative information distributed across the full forward pass. We introduce HalluTracer, a detection framework that reads and aggregates truthfulness evidence across every layer of the forward pass before the model emits any answer token. A geometric analysis reveals that the per-layer signals are weakly correlated, so that simple depth averaging suppresses layer-specific noise and captures nearly all linearly accessible information. Across six open-source language models and five hallucination benchmarks, HalluTracer consistently outperforms matched white-box baselines, with gains ranging from one to fourteen points. Collectively, our work recasts hallucination detection from a layer-selection problem into a depth-aggregation problem governed by the geometric sparsity of the truthfulness signal. △ Less

Submitted 17 August, 2026; originally announced August 2026.

MicroVerse: An Instrument for Measuring Self-Authored Identity Drift in Long-Horizon Multi-Agent Language-Model Simulations

Authors: Sky Ng, Brihi Joshi, Ishan Gupta, Shirley Huang, Zonglin Di, Yun Shen, Qianfeng Wen, Yifan Simon Liu, Ruoqi Gao, Yilan, Fan, Zhiwei Zhang, Muhammad Ahmed Mohsin, Yucheng Lu, Xiaoyi Liu, Heming Liu, Qianyu Zhu, Hanwen Xing, Zhengyang Shan, My Chiffon Nguyen, Guanghui Min, Jianheng, Hou, Yunze, Xiao , et al. (25 additional authors not shown)

Abstract: Long-horizon, multi-agent language model (LM) simulations are widely proposed for studying social behavior, yet instruments to measure whether persona-conditioned agents maintain identity fidelity under sustained pressure are lacking. We present MicroVerse, a behavioral-science instrument that measures identity drift in generative agents. Agents carry an immutable "soul file" (core values, moral b… ▽ More Long-horizon, multi-agent language model (LM) simulations are widely proposed for studying social behavior, yet instruments to measure whether persona-conditioned agents maintain identity fidelity under sustained pressure are lacking. We present MicroVerse, a behavioral-science instrument that measures identity drift in generative agents. Agents carry an immutable "soul file" (core values, moral boundaries, personality, goals) and inhabit a resource-scarce 50 x 50 environment where water is a non-respawning survival constraint. Scarcity is operationalized via a per-tick existence-cost gradient. The eight-verb action space maps directly to moral boundaries (trade, talk, attack, scavenge). Using a three-layer memory architecture, agents periodically revise a mutable current identity against their immutable original soul via importance-triggered reflection. To mitigate survivor bias, MicroVerse decouples measurement from behavior using uniform longitudinal engine snapshots every N ticks alongside a forced-end snapshot of all living and dead agents. Identity drift is scored offline using a paraphrase-aware, value-anchored, multi-register diff rather than raw cosine similarity. We evaluate the instrument via a controlled seed run (n = 25) and a reflection-threshold sweep (thresholds {40, 80, 150}) to determine if drift dynamics are gate artifacts or threshold-robust properties. We report two primary findings: (1) Anti-self-deception emerges unprompted as the single largest semantic category of identity modification (27 of 111 added boundaries, 24%). (2) The system is threshold-robust; lower gates accelerate and increase revision frequency but preserve drift direction. All empirical results are strictly preliminary existence proofs and effect shapes (one model, one seed per arm, n = 25) rather than statistical significance claims. △ Less

Submitted 16 August, 2026; originally announced August 2026.

PersonaEval: Persona-Based User Simulation for Evaluating Interactive Applications

Authors: Yifan Simon Liu, Qianfeng Wen, Yilan Fan, Shirley Huang, Ruoqi Gao, Jianheng Hou, Muhammad Ahmed Mohsin, Zonglin Di, Brihi Joshi, Xincheng Tan, Yucheng Lu, Xiaoyi Liu, Heming Liu, Hanwen Xing, Guanghui Min, Zhengyang Shan, My Chiffon Nguyen, Ishan Gupta, Yunze Xiao, Hannah Collison, Jintao Huang, Jiatong Li, Sankalp Jajee, Yunhan Zhao, Bing Hu , et al. (18 additional authors not shown)

Abstract: Real user studies are important for understanding how people interact with systems under test or already deployed. In practice, however, they are often costly, time-consuming, and difficult to scale. To address these challenges, we introduce PersonaEval, a persona-based user simulation framework that approximates real-user behavior across diverse interactive settings. PersonaEval connects simulate… ▽ More Real user studies are important for understanding how people interact with systems under test or already deployed. In practice, however, they are often costly, time-consuming, and difficult to scale. To address these challenges, we introduce PersonaEval, a persona-based user simulation framework that approximates real-user behavior across diverse interactive settings. PersonaEval connects simulated users drawn from existing persona datasets to task-specific application interfaces and collects the interaction trajectories and outcomes. PersonaEval provides a plug-and-play evaluation workflow in which the application being evaluated can be easily changed. In this demo, we present PersonaEval on three forms of interactive applications: surveys, chatbots, and web applications. Together, these examples show that PersonaEval can support repeatable, parallelizable, and scalable evaluation across different interaction settings, while producing user-oriented feedback and task-specific behavior. △ Less

Submitted 16 August, 2026; originally announced August 2026.

TimeSage-EV: A Live Benchmark for Agentic Time Series Analysis in Evolving Environments

Authors: Qingren Yao, Yaxuan Kong, Yuqi Nie, Yichen Li, Stefan Zohren, Anna Vettoruzzo, Qingsong Wen, Ming Jin, Joaquin Vanschoren

Abstract: Time series analysis in high-stakes domains relies on recurring data releases, where new observations can alter the evidence base and the validity of later conclusions. Existing time series QA benchmarks mostly rely on fixed snapshots, leaving temporal validity and cutoff-aware evidence use unevaluated. We introduce TimeSage-EV, a live benchmark for agentic time series analysis in evolving environ… ▽ More Time series analysis in high-stakes domains relies on recurring data releases, where new observations can alter the evidence base and the validity of later conclusions. Existing time series QA benchmarks mostly rely on fixed snapshots, leaving temporal validity and cutoff-aware evidence use unevaluated. We introduce TimeSage-EV, a live benchmark for agentic time series analysis in evolving environments. It tracks 60 real institutional scenarios across 6 domains, comprising 1,485 scenario-period QA pairs from Feb 2023 to May 2026 and spanning monthly, weekly, daily, and irregular release cadences. At each period, large language model (LLM) agents receive time series data and source reports, while the withheld target release provides ground truth. TimeSage-EV evaluates state identification, data summarization, and outlook reasoning. Experiments with frontier LLM agents and TimeSage-1.0, a novel self-evolving agent with a reusable analytical skill library, reveal significant performance gaps across model tiers and recurring failures in temporal validity, exogenous context use, and adaptation. We release TimeSage-EV as a research resource with monthly updates, code, a leaderboard, and failure-mode analyses. △ Less

Submitted 14 August, 2026; originally announced August 2026.

Forecast Collapse in Time-Series Foundation Models

Authors: Shu Wan, Miles Ma, Hank Zhu, Guangqi Liu, Stephen Wang, Qingsong Wen, Huan Liu

Abstract: When forecasting hourly returns for 1,000 US equities, we observe an unexpected phenomenon: predictions become nearly flat and show poor stock ranking, as measured by cross-sectional correlation. We call this forecast collapse. Surprisingly, the phenomenon largely disappears when forecasting trading volume under the same setting. We investigate forecast collapse across time-series foundation model… ▽ More When forecasting hourly returns for 1,000 US equities, we observe an unexpected phenomenon: predictions become nearly flat and show poor stock ranking, as measured by cross-sectional correlation. We call this forecast collapse. Surprisingly, the phenomenon largely disappears when forecasting trading volume under the same setting. We investigate forecast collapse across time-series foundation models (TSFMs), twelve deep-learning forecasting models, and 97 public benchmark configurations, and find that it is closely tied to target predictability. We identify two distinct reasons behind it: low predictability limits the amplitude of calibrated point forecasts, while per-series objectives leave cross-series structure unidentified. These findings reveal a calibration-ranking tradeoff: optimizing squared error leads to flat predictions, whereas directly optimizing cross-sectional correlation improves ranking but can inflate forecast amplitude by more than an order of magnitude. To address this tradeoff, we introduce CalibRank, a simple objective that balances calibration and ranking. On Finance1K, CalibRank nearly triples cross-sectional correlation while keeping amplitude close to the target, and improves correlation on all tested models. Our results reveal a blind spot in conventional time-series evaluation: per-series metrics can hide failures in cross-series structure needed by downstream decisions. △ Less

Submitted 14 August, 2026; originally announced August 2026.

Comments: 27 pages, 3 figures, 5 tables. Dataset: https://huggingface.co/datasets/abel-lab/finance1k

An Adaptive Longitudinal Platooning Design Based On Concurrent Learning

Authors: Qiuhao Wen, Di Liu, Jiwei Wang, Simone Baldi

Abstract: This work proposes a new adaptive longitudinal platooning strategy in the framework of concurrent learning. Adaptive refers to vehicles facing uncertainty in powertrain parameters via on-line estimation; concurrent learning refers to using both current and past data in the estimation. The proposed platooning strategy advances existing ones since convergence to the true powertrain parameters is gua… ▽ More This work proposes a new adaptive longitudinal platooning strategy in the framework of concurrent learning. Adaptive refers to vehicles facing uncertainty in powertrain parameters via on-line estimation; concurrent learning refers to using both current and past data in the estimation. The proposed platooning strategy advances existing ones since convergence to the true powertrain parameters is guaranteed without imposing persistence of excitation on the vehicle behavior: it suffices the presence of a single non-zero data sample. Meanwhile, the concurrent learning proof we give advances existing ones since it takes into account an extra unknown gain in the error dynamics. △ Less

Submitted 7 August, 2026; originally announced August 2026.

Journal ref: IEEE Control Systems Letters, vol. 8, pp. 303-308, 2024

Correct Online Estimation of the Powertrain Time Constants in Adaptive Vehicular Platooning

Authors: Qiuhao Wen, Simone Baldi, Jiwei Wang, Wenwu Yu, Di Liu

Abstract: In longitudinal platooning, some key sources of uncertainty are the powertrain time constants of the vehicles. Because such time constants appear in the input matrix of the platooning dynamics, their correct estimation is either impractical with methods requiring persistence of excitation, or impossible with methods requiring the input matrix to be known. This work proposes a novel adaptive longit… ▽ More In longitudinal platooning, some key sources of uncertainty are the powertrain time constants of the vehicles. Because such time constants appear in the input matrix of the platooning dynamics, their correct estimation is either impractical with methods requiring persistence of excitation, or impossible with methods requiring the input matrix to be known. This work proposes a novel adaptive longitudinal platooning method with correct estimation of the powertrain time constants. To achieve correct estimation, the composite adaptive control framework and its stability analysis are suitably modified to handle the time constant uncertainty in the design of the adaptive law. The result is a platooning protocol that guarantees convergence of the estimated time constants to their true values without the need for persistence of excitation: it is sufficient the derivative of the acceleration to be nonzero over a possibly short transient, an extremely relaxed excitation condition. Comparisons with state-of-the-art platooning solutions reveal advantages such as no required measurements of acceleration derivative nor collection of past data. The robustness and practicality of the proposed design is also verified with CarSim-based platooning experiments. △ Less

Submitted 7 August, 2026; originally announced August 2026.

Journal ref: IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 56, no. 1, pp. 279-291, Jan. 2026

SR-JEPA: Learning Predictive Latent State in 3D Scenes

Authors: Zihan Zhou, Qifu Wen, Xi Zeng

Abstract: Joint-embedding predictive architectures learn by predicting latent representations of missing observations, yet many masked JEPAs are evaluated primarily through the encoders they produce. We ask what a trained predictive pathway itself infers when an entire entity is absent from a native 3D scene. We introduce SR-JEPA, a point-native JEPA for scene-scale point clouds whose original frozen predic… ▽ More Joint-embedding predictive architectures learn by predicting latent representations of missing observations, yet many masked JEPAs are evaluated primarily through the encoders they produce. We ask what a trained predictive pathway itself infers when an entire entity is absent from a native 3D scene. We introduce SR-JEPA, a point-native JEPA for scene-scale point clouds whose original frozen predictive pathway can be queried at a supplied location. At evaluation, every point of one object is removed before encoding and replaced by the same shape-free 32-point query at its centroid. Training uses only self-contained 3D EMA targets: no reconstruction, semantic labels, language, or lifted 2D features. On 5,953 held-out ARKitScenes objects, the imputed latent reaches 43.13% semantic-identity macro accuracy, 22.18 points above the strongest floor. Randomizing the prediction path removes 9.78 points, while substituting matched donor context removes 21.98 points. On 8,570 Sr3D support pairs, the full latent reaches 41.15 AP; identity decoded from the missing-object latent, combined with anchor identity and geometry, reaches 39.37 AP, leaving an unresolved 1.78-point residual. These results reveal a queryable, compositional 3D predictive state: the model completes context-dependent entity content, which downstream computation combines with metric geometry. △ Less

Submitted 6 August, 2026; originally announced August 2026.

Comments: 17 pages, 5 figures, 9 tables

ACM Class: I.2.10; I.2.6; I.4.8

MatrAIx: Simulating the World with 8.3 Billion Persona Agents

Authors: Xiaomin Li, Yuexing Hao, Jianheng Hou, Jintao Huang, Qianfeng Wen, Shirley Huang, Yifan Liu, Xiaoyi Liu, Yilan Fan, Yijun Wang, Koutian Wu, Ruoqi Gao, Muhammad Ahmed Mohsin, Jing Tang, Brihi Joshi, Heming Liu, Zheyuan Deng, Zonglin Di, Sankalp Jajee, Jiuyao Lu, Zhiwei Zhang, Saksham Kapoor, Ishan Gupta, Yunhan Zhao, Chanwoo Park , et al. (68 additional authors not shown)

Abstract: Human evaluation of AI systems and digital products is costly, slow, and difficult to scale. Offline evaluations are more scalable but often abstract away human diversity and interactive behavior. We therefore introduce MatrAIx, a population-scale simulated-user evaluation infrastructure for testing AI systems and digital products with heterogeneous users. MatrAIx has three core components: First,… ▽ More Human evaluation of AI systems and digital products is costly, slow, and difficult to scale. Offline evaluations are more scalable but often abstract away human diversity and interactive behavior. We therefore introduce MatrAIx, a population-scale simulated-user evaluation infrastructure for testing AI systems and digital products with heterogeneous users. MatrAIx has three core components: First, Persona 8B contains 8.3 billion persona records represented by 1,290 categorical dimensions. Records are either sampled from a dependency graph that preserves correlated attributes or derived from human-authored profiles. We release a quality-filtered coreset of approximately 1 million personas, comprising 599,847 human-grounded and 400,000 synthetic records. Second, the MatrAIx Playground provides four environments in which diverse users evaluate and interact with digital products: Survey, AI Chatbot, Web, and App. Third, MatrAIx provides 1,010 application tasks spanning more than 25 domains, including Commerce, Software, Finance, and Healthcare. We conducted 18,189 evaluation trials across eight representative tasks. Persona agents were powered by three LLMs: Claude Opus 4.8, GPT 5.5, and Claude Haiku 4.5. The resulting feedback captures how decisions and preferences vary across persona backgrounds, including hesitation after a price increase, willingness to continue after an AI assistant fails, and latency tolerance. We conducted two main validation studies: First, a 400-trial controlled study evaluated persona adherence across ten behavioral attributes and all four environments. The declared behavior was expressed or correctly suppressed in 366 trials (91.5%). Second, human and LLM judges evaluated the extraction quality of human-grounded personas. Overall, MatrAIx provides an end-to-end infrastructure for evaluating AI systems and digital products with diverse simulated human users. △ Less

Submitted 4 August, 2026; originally announced August 2026.

Comments: Project website: https://matraix.ai

CoWAM: Coordination Contracts for Selective Policy Intervention with WAMs

Authors: Shuaijun Liu, Qifu Wen, Shuyang Hao, Qi Luo, Chenglong Zhang, Feiyang You, Chengyu Wu, Ningxin Su

Abstract: World Action Models (WAMs) augment robot policies with action-conditioned predicted futures, but a plausible future alone does not justify changing the action that a bimanual policy would execute. We present CoWAM, a selective intervention layer that expresses synchronization, role compatibility, and collision convergence as coordination contracts. Each contract combines typed admissibility checks… ▽ More World Action Models (WAMs) augment robot policies with action-conditioned predicted futures, but a plausible future alone does not justify changing the action that a bimanual policy would execute. We present CoWAM, a selective intervention layer that expresses synchronization, role compatibility, and collision convergence as coordination contracts. Each contract combines typed admissibility checks with event-conditioned verification and calibrated intervention gates. CoWAM preserves the nominal action unless an alternative satisfies every active obligation and provides a clear, low-risk improvement; when the nominal action is also inadmissible, it invokes a predefined abstention fallback. To separate selector quality from proposal quality, all methods operate on identical candidate pools and commit their decisions before shared oracle labeling. Across eight simulated bimanual tasks, CoWAM improves coordination-valid selection by 16.7 percentage points over the contract-only variant and raises closed-loop success by 9.6 percentage points over the strongest selective baseline, while keeping harmful interventions below 1%. Together, these results establish coordination contracts as an effective interface for conservative policy intervention with predicted world-action evidence across coordination-rich bimanual tasks. △ Less

Submitted 3 August, 2026; originally announced August 2026.

Adapting Embedding Models for Agent Capability Retrieval

Authors: Tingwei Chen, Yunxiao Shi, Zhengdong Chu, Qingsong Wen, Min Xu

Abstract: Open agent marketplaces list native agents, tool bundles, and reusable skill packages in the same search interface, yet practitioners still have little guidance on how to retrieve across this mixed catalog. We study whether off-the-shelf retrieval models, trained for general text retrieval, can be adapted to match user queries to executable agent capabilities, and whether the learned signal transf… ▽ More Open agent marketplaces list native agents, tool bundles, and reusable skill packages in the same search interface, yet practitioners still have little guidance on how to retrieve across this mixed catalog. We study whether off-the-shelf retrieval models, trained for general text retrieval, can be adapted to match user queries to executable agent capabilities, and whether the learned signal transfers beyond the benchmark used for tuning. We fine-tune three open retrieval backbones, BGE-base, KaLM-v1.5, and EasyRec, on AgentSelect, which represents marketplace-visible units as capability profiles derived from public metadata, and test transfer on two catalogs not seen during training: MuleRun native agents and a ClawHub benchmark of 50 skills with 1,000 queries. Adaptation helps on both catalogs. Code and data will be released upon publication. △ Less

Submitted 19 July, 2026; originally announced July 2026.

Comments: Accepted for oral presentation at the AgentSearch Workshop, SIGIR 2026

M3F-UAV: A Missing-Modality Multimodal Foundation Model for Low-Altitude Wireless Sensing

Authors: Pengxuan Gao, Kai Ying, Botao Wu, Jianhua Mo, Qingsong Wen

Abstract: Low-altitude unmanned aerial vehicles (UAVs) are emerging as key platforms for wireless intelligence tasks. However, practical low-altitude wireless systems usually operate in complex urban environments, where visual occlusion, sparse geometric observations, multipath propagation, and sensor failures may degrade the reliability of single-modality models. To address these challenges, this paper pro… ▽ More Low-altitude unmanned aerial vehicles (UAVs) are emerging as key platforms for wireless intelligence tasks. However, practical low-altitude wireless systems usually operate in complex urban environments, where visual occlusion, sparse geometric observations, multipath propagation, and sensor failures may degrade the reliability of single-modality models. To address these challenges, this paper proposes M3F-UAV, a missing-modality multimodal foundation model for low-altitude wireless sensing. The proposed framework learns a unified multimodal representation from visual, geometric, and wireless observations. Specifically, modality-specific pretrained feature extractors are adopted for RGB/depth images, LiDAR point clouds, and CSI matrices, respectively. Through cross-modal fusion and missing-modality-aware pretraining with feature-level masked reconstruction and UAV localization objectives, M3F-UAV can extract fixed-size features from different modality combinations and adapt them to downstream low-altitude wireless tasks with lightweight task heads. Experiments on the LAMBDA dataset show that M3F-UAV outperforms single-modality baselines and maintains robust performance under missing-modality settings. △ Less

Submitted 15 July, 2026; originally announced July 2026.

Continuum modeling of fluidic and elastic flow during growth-driven wound closure in partial-EMT cell monolayers

Authors: Chaozhen Wei, Han Jiang, Yifan Gu, Nonthakorn Olaranont, Pengbo Wang, Qi Wen, Yubing Sun, Min Wu

Abstract: Large-scale circular gap closure occurs over a time scale on which cell growth and proliferation become important. Growth is the main driver of the closing process, while cell dynamics such as elongation and intercalation reflect elastic and fluidic contributions to tissue deformation. We develop a novel fluidized growth-elasticity framework as a nonlinear analogue of a Maxwell fluid with growth.… ▽ More Large-scale circular gap closure occurs over a time scale on which cell growth and proliferation become important. Growth is the main driver of the closing process, while cell dynamics such as elongation and intercalation reflect elastic and fluidic contributions to tissue deformation. We develop a novel fluidized growth-elasticity framework as a nonlinear analogue of a Maxwell fluid with growth. The framework decomposes the experimentally observable strain rate into the additive sum of the growth, elastic, and fluidic strain rates, thus enabling the separate quantification of these contributions from tissue kinematics and allowing the roles of tissue elasticity and fluidity (the inverse of viscosity) to be characterized. We apply the model to large circular gaps ($\sim$1.7 mm in diameter) in confluent monolayers of mouse embryonic epicardial cells (MEC1) under two conditions, without and with TGF-$β$ treatment. We show that both tissue fluidity and the elastic properties associated with fiber reinforcement are critical for reproducing the closure kinematics. Specifically, we predict that the treated condition has lower fluidity, associated with a lower fluidic deformation rate and a higher elastic deformation rate than the untreated condition, in agreement with the experimental observations. △ Less

Submitted 14 August, 2026; v1 submitted 7 July, 2026; originally announced July 2026.

Comments: 36 pages, 14 figues

MSC Class: 35Q74; 92Cxx; 74L15; 74Bxx

Holography and Kinematic Space for Gravitational Sub-regions in AdS

Authors: Debarshi Basu, Qiang Wen

Abstract: It is well-known in integral geometry that a maximally symmetric Riemannian manifold, such as a static slice of vacuum AdS spacetime, can be perfectly covered by the geodesics in the Kinematic space, which we call the partial-entanglement-entropy (PEE) threads. In this context, the area of a codimension-one surface in the manifold can be computed by counting its intersections with the PEE threads,… ▽ More It is well-known in integral geometry that a maximally symmetric Riemannian manifold, such as a static slice of vacuum AdS spacetime, can be perfectly covered by the geodesics in the Kinematic space, which we call the partial-entanglement-entropy (PEE) threads. In this context, the area of a codimension-one surface in the manifold can be computed by counting its intersections with the PEE threads, which is the celebrated Crofton formula. In this paper, we analyze the Kinematic space for a generic subregion in vacuum AdS space, and propose that the PEE threads emanate from a co-dimension one surface can perfectly cover a subregion in the manifold. Furthermore, we build holographic tensor network models on the network of the PEE threads confined in a subregion, thereby providing a concrete framework that realizes the surface-state correspondence and the generalized entanglement wedges for gravitational subregions. △ Less

Submitted 29 July, 2026; v1 submitted 29 June, 2026; originally announced June 2026.

Comments: 25 pages, 12 figures, comments are welcome; v2 references added.;v3,minor revision on the presentation

Comb-enabled spectral-domain image transport through perturbation-prone multimode fibers

Authors: Maohan Li, Zijian Wang, Bowen Sun, Zhuoren Wan, Xiangze Ma, Xiuxiu Zhang, Yuan Chen, Mei Yang, Qi Wen, Zhaoyang Wen, Ming Yan, Heping Zeng

Abstract: Multimode fibers (MMFs) offer a compact platform for imaging, sensing, and information transport, but their practical deployment is hindered by sensitivity to fiber perturbations, which alter modal coupling and invalidate conventional speckle-based calibrations. Here, we demonstrate perturbation-resilient image transport through MMFs by combining image-to-spectrum encoding with dual-comb spectrosc… ▽ More Multimode fibers (MMFs) offer a compact platform for imaging, sensing, and information transport, but their practical deployment is hindered by sensitivity to fiber perturbations, which alter modal coupling and invalidate conventional speckle-based calibrations. Here, we demonstrate perturbation-resilient image transport through MMFs by combining image-to-spectrum encoding with dual-comb spectroscopy. Two-dimensional images are converted into comb-line-resolved spectral signatures before fiber transmission, allowing spatial information to be carried in the spectral domain rather than in the output speckle field. After propagation, dual-comb heterodyne detection maps the encoded spectrum into the radio-frequency domain, enabling massively parallel spectral readout with a single photodetector. Neural-network-assisted compressive reconstruction further enables high-fidelity imaging from sparse, noisy, and spectrally aliased measurements. Our approach achieves Pearson correlation coefficients exceeding 0.9 under strong fiber perturbations and supports frame rates up to 2.5 MHz, allowing the observation of transient switching dynamics in a digital micromirror device. These results establish a powerful tool for robust, real-time image transport through flexible MMFs, with potential applications in remote sensing and fiber-based optical instrumentation. △ Less

Submitted 29 June, 2026; originally announced June 2026.

Comments: 20 pages,4 figures

SafeGEO: Understanding Generative Engine Optimization Risks in Recommendation Agents

Authors: Qianfeng Wen, Yifan Simon Liu, Xin Liu, Difan Jiao, Blair Yang, Junda Wu, Zhenwei Tang

Abstract: Generative Engine Optimization (GEO) lets content owners rewrite web content to increase their visibility in generative systems. In recommendation agents, this creates a risk that seller-controlled sources make flawed products appear better supported than they are. We study this risk by asking whether recommendation agents preserve utility-aligned decisions when seller-controlled sources are rewri… ▽ More Generative Engine Optimization (GEO) lets content owners rewrite web content to increase their visibility in generative systems. In recommendation agents, this creates a risk that seller-controlled sources make flawed products appear better supported than they are. We study this risk by asking whether recommendation agents preserve utility-aligned decisions when seller-controlled sources are rewritten for GEO. To make this question measurable, we construct SafeGEO, an evaluation suite with 22 GEO attack variants across 600 recommendation cases. We empirically show that GEO attacks can promote flawed target products. On average, they increase the rate at which such flawed products enter the recommendation set by up to 83.2%. We further study whether agent-side design choices can mitigate this risk and show that simple defenses, including defensive prompting and structured evidence checks, reduce harmful target promotion by up to 39.2%. These gains are substantial but do not restore the no-GEO performance, showing that GEO remains a serious risk despite developer-side mitigation. △ Less

Submitted 8 June, 2026; originally announced June 2026.

Comments: 41 pages,23 figures

FedUP: One-Shot Federated Unlearning via Centroid-Guided Plug-in Filters

Authors: Feihong Nan, Zhengyi Zhong, Pan Wang, Weidong Bao, Xiongtao Zhang, Quan Wen, Ji Wang

Abstract: Federated unlearning (FU) is critical for complying with legal mandates like the right to be forgotten in decentralized systems, yet current methods face a persistent dilemma between non-target knowledge loss and high request latency. To resolve these issues, we propose FedUP, a one-shot federated unlearning framework utilizing lightweight pluggable filters that act as a "knowledge funnel" to scre… ▽ More Federated unlearning (FU) is critical for complying with legal mandates like the right to be forgotten in decentralized systems, yet current methods face a persistent dilemma between non-target knowledge loss and high request latency. To resolve these issues, we propose FedUP, a one-shot federated unlearning framework utilizing lightweight pluggable filters that act as a "knowledge funnel" to screen out target data while preserving original model performance. By freezing original model parameters and training filters at the server side using differentially private (DP)-protected class centroid samples, FedUP bypasses the need for multi-round client-server communication and complex retraining, reducing unlearning latency from minutes to mere seconds. Additionally, the framework's pluggable architecture ensures inherent reversibility, enabling the seamless restoration of forgotten knowledge by simply removing the filters. Extensive experiments on diverse image and text tasks demonstrate that FedUP effectively reduces non-target knowledge loss and achieves superior unlearning precision and efficiency across various scenarios. Code is available at: https://github.com/suows/FedUP-code. △ Less

Submitted 24 June, 2026; v1 submitted 22 June, 2026; originally announced June 2026.

Comments: Accepted by IJCAI 2026

Calibration Is Not Control: Why LLM-Agent Oversight Needs Intervention

Authors: Chubin Zhang, Zhenglin Wan, Xingrui Yu, Jingxuan Wu, Qi Wen, Pengfei Zhou, Wangbo Zhao, Ivor Tsang

Abstract: Runtime oversight for LLM agents is commonly framed as scalar risk prediction: estimate failure likelihood, confidence, or uncertainty, then intervene once the score crosses a threshold. We argue that this framing targets the wrong object for control. The relevant question is not how likely the agent is to fail if it continues, but whether an available intervention would improve the outcome. Two t… ▽ More Runtime oversight for LLM agents is commonly framed as scalar risk prediction: estimate failure likelihood, confidence, or uncertainty, then intervene once the score crosses a threshold. We argue that this framing targets the wrong object for control. The relevant question is not how likely the agent is to fail if it continues, but whether an available intervention would improve the outcome. Two trajectory prefixes can have the same risk estimate while requiring different actions, because one remains recoverable and the other does not. We formalize this mismatch as target error and identify intervention advantage, the expected utility gain from intervening rather than continuing, as the decision object for oversight. To measure this mismatch, we introduce prefix branching, a same-prefix counterfactual protocol that executes candidate actions from identical trajectory states. Across four benchmarks, action-conditioned control yields regime-dependent gains over scalar routing. In a calibration decomposition, recalibrating the same scalar score improves prediction metrics but leaves control regret unchanged, showing that calibration alone does not repair target error. A simple prefix-only action-conditioned controller substantially reduces regret in the strongest interactive regime, from 0.506 to 0.110 on ALFWorld. Gains shrink when interventions are weak or when scalar routing already preserves intervention-relevant information. These results suggest that LLM-agent oversight should move from calibrated risk scoring toward action-conditioned value estimation. △ Less

Submitted 19 June, 2026; originally announced June 2026.

Comments: 29 pages

From Memorization to Creation: Evaluating the Cognitive Depth of LLM-Generated Educational Questions

Authors: Xiaolong Wang, Zhe Zhao, Song Lai, Chaoli Zhang, Zijie Geng, Yu Tong, Ye Wei, Qingsong Wen

Abstract: While LLMs show promise in automating educational content creation, their ability to generate questions that stimulate higher-order thinking remains understudied. This work evaluates six widely-used LLMs through a Bloom's Taxonomy lens, focusing on their capacity to transcend rote memorization and achieve cognitive leaps. Using a hybrid human--AI evaluation protocol, we generate and analyze 20{,}7… ▽ More While LLMs show promise in automating educational content creation, their ability to generate questions that stimulate higher-order thinking remains understudied. This work evaluates six widely-used LLMs through a Bloom's Taxonomy lens, focusing on their capacity to transcend rote memorization and achieve cognitive leaps. Using a hybrid human--AI evaluation protocol, we generate and analyze 20{,}700 questions across computer science, K--12 math, and social-science domains. Key contributions include: (1) a fine-grained prompting strategy that reduces question repetitiveness by 24.45\% for Qwen2.5-7B-Instruct, and increases the proportion of higher-order cognitive level outputs by 11.53\% for InternLM3-8B-Instruct; (2) quantitative metrics for cognitive shift intensity (CogShift) and category drift, revealing InternLM3's superior performance in multi-level transitions; (3) an interpretability analysis revealing metric-level correlations that enhance the transparency of Chain-of-Thought prompting. Our findings highlight the importance of cognitive-aware prompt design and provide benchmarks for deploying LLMs in personalized learning systems. △ Less

Submitted 5 May, 2026; originally announced June 2026.

Comments: Accepted by KDD 2026

Journal ref: KDD 2026

Oxidation-induced ultrafast spin-to-orbital conversion at heavy-metal interfaces

Authors: Xiaoxue Zeng, Tianyi Zhang, Yaokai Niu, Qiye Wen, Zhiyong Zhong, Zhi-Min Liao, Peng Yan, Xiufeng Han, Lichuan Jin

Abstract: Oxidation engineering provides a route to control orbital degrees of freedom, yet its role in spin-to-orbital conversion remains largely unexplored. Here, we report an efficient spin-to-orbital conversion mechanism driven by interfacial oxidation at heavy-metal interfaces. In W/Co/SiO2 heterostructures, terahertz emission exhibits a time delay that scales linearly with the W thickness, identifying… ▽ More Oxidation engineering provides a route to control orbital degrees of freedom, yet its role in spin-to-orbital conversion remains largely unexplored. Here, we report an efficient spin-to-orbital conversion mechanism driven by interfacial oxidation at heavy-metal interfaces. In W/Co/SiO2 heterostructures, terahertz emission exhibits a time delay that scales linearly with the W thickness, identifying orbital-current transport as the dominant origin. The emission amplitude is approximately three times larger than that of Co/Pt bilayers, indicating highly efficient conversion from spin to orbital angular momentum. Systematic variation of Co thickness, stoichiometry, and interface configuration reveals that the effect originates from oxidation of the W layer at the W/Co interface, which modulates the interfacial orbital texture. We further show that this mechanism is generic across different heavy metals and scales with their spin-orbit coupling strength. These results establish oxidation as an effective handle to engineer spin-to-orbital conversion and provide a general route toward orbitronic terahertz emitters. △ Less

Submitted 13 June, 2026; originally announced June 2026.

Comments: 18 pages, 4 figures

PAC-Bayesian Adversarially Robust Generalization for Message Passing Graph Neural Networks: A Sensitivity Analysis

Authors: Ziling Liang, Xinping Yi, Qingsong Wen, Shi Jin

Abstract: Whilst the vulnerability of graph neural networks (GNNs) to adversarial attacks poses a critical threat to graph representation learning, the understanding of the robust generalization behavior remains a fundamental challenge in the adversarial setting. Recently, PAC-Bayesian margin-based generalization analysis substantially advances this line of research by providing a flexible and data-dependen… ▽ More Whilst the vulnerability of graph neural networks (GNNs) to adversarial attacks poses a critical threat to graph representation learning, the understanding of the robust generalization behavior remains a fundamental challenge in the adversarial setting. Recently, PAC-Bayesian margin-based generalization analysis substantially advances this line of research by providing a flexible and data-dependent analytical framework. However, existing robust analyses often rely on isotropic Gaussian posteriors and control weight perturbations in the full parameter space, which limits the ability to capture heterogeneous parameter sensitivity yet hinges on hidden-width-dependent complexity terms, resulting in not-tight-enough generalization bounds. In this paper, we extend a recently proposed sensitivity-aware PAC-Bayesian framework from deep neural networks to message passing GNNs (MPGNNs) and derive a tighter robust generalization bound in the adversarial setting. Specifically, we first quantify how sensitive the perturbations across different parameter blocks are to the network outputs by deriving the output Jacobians with respect to the weight parameters. Exploiting the fact that these Jacobian matrices have rank at most $K$ in $K$-class graph classification, we then construct Jacobian-aligned sensitivity matrices and use anisotropic Gaussian posteriors with optimized covariances to upper bound the KL divergence in a tight way. Notably, by refining the spectral-norm dependence on the learned weights and reducing the leading dimension factor from hidden-width-dependent terms to the number of classes $K$, our analysis yields much tighter robust generalization guarantees for MPGNNs, thereby guiding their designs to enhance adversarial robustness. △ Less

Submitted 4 June, 2026; originally announced June 2026.

TimeSage-MT: A Multi-Turn Benchmark for Evaluating Agentic Time Series Reasoning

Authors: Yaxuan Kong, Qingren Yao, Yuqi Nie, Yichen Li, Yilei Shao, Stefan Zohren, Anna Vettoruzzo, Joaquin Vanschoren, Ming Jin, Qingsong Wen

Abstract: Time series data inform critical decisions across many real-world domains. While large language model (LLM) agents can analyze data through natural language and tools, it remains unclear whether they can conduct reliable time series analysis across multi-turn conversations. Existing benchmarks focus on single-step tasks such as forecasting and anomaly detection, overlooking practical workflows whe… ▽ More Time series data inform critical decisions across many real-world domains. While large language model (LLM) agents can analyze data through natural language and tools, it remains unclear whether they can conduct reliable time series analysis across multi-turn conversations. Existing benchmarks focus on single-step tasks such as forecasting and anomaly detection, overlooking practical workflows where user goals evolve, agents must build on prior analyses, and conclusions emerge from accumulated evidence. In this work, we introduce TimeSage-MT, a multi-turn benchmark for agentic time series reasoning with 240 tasks and 2,680 dialogue turns across 8 real-world domains, spanning basic exploration to decision-oriented analysis. TimeSage-MT is built through a reproducible pipeline that converts real-world time series data into multi-turn conversations with verifiable answers. It provides a unified evaluation protocol and public leaderboard for comparing time series agentic systems. To demonstrate the benchmark's utility, we evaluate frontier LLMs alongside TimeSage, a novel structured agent equipped with a comprehensive time series skill library. The results show sharp performance drops on decision-oriented tasks, driven by failures in memory, uncertainty handling, and domain-based decision making. TimeSage-MT exposes critical gaps in current agentic reasoning and provides a rigorous foundation for future development. △ Less

Submitted 31 May, 2026; originally announced June 2026.

SkillBrew: Multi-Objective Curation of Skill Banks for LLM Agents

Authors: Wentao Hu, Zhendong Chu, Yiming Zhang, Junda Wu, Ming Jin, Xiangyu Zhao, Yilei Shao, Yanfeng Wang, Qingsong Wen

Abstract: Retrieval-augmented LLM agents increasingly rely on curated skill banks: collections of reusable textual principles that guide decision making on complex tasks. Existing approaches typically expand these banks in an append-only fashion, continuously adding new skills without removing redundant, outdated, or harmful ones, resulting in inefficient and poorly curated repositories. In this paper, we f… ▽ More Retrieval-augmented LLM agents increasingly rely on curated skill banks: collections of reusable textual principles that guide decision making on complex tasks. Existing approaches typically expand these banks in an append-only fashion, continuously adding new skills without removing redundant, outdated, or harmful ones, resulting in inefficient and poorly curated repositories. In this paper, we formulate the skill bank curation as a constrained multi-objective problem: a desirable bank must be useful for the agent, diverse in its content, and provide good coverage of the query distribution. To this end, we introduce SkillBrew, a multi-objective curation framework that formalizes skill bank curation as Pareto-aware optimization under a utility constraint, and solves it via a bi-level propose-then-verify loop. We evaluate our approach on two public benchmarks. Our findings suggest that treating skill banks as objects of principled curation, rather than ever-growing append-only logs, is an important step toward building self-improving LLM agents. △ Less

Submitted 28 May, 2026; originally announced May 2026.

Comments: 16 pages. Preprint. Under review

Physics-Aware 3D Gaussian Editing for Driving Scene Generation

Authors: Feng Zhou, Jian Zhang, Yuhang Sun, He Wang, Qiong Wen, Debao Kong, Tieru Wu, Rui Ma

Abstract: 3D Gaussian Splatting (3DGS) has shown great potential in autonomous driving simulation and data generation, enabling photorealistic reconstruction and flexible scene manipulation. However, existing 3DGS scene editing methods have limited support for road geometry editing (e.g., inserting speed humps or sunken roads), and generally do not couple such edits with plausible vehicle-road interaction d… ▽ More 3D Gaussian Splatting (3DGS) has shown great potential in autonomous driving simulation and data generation, enabling photorealistic reconstruction and flexible scene manipulation. However, existing 3DGS scene editing methods have limited support for road geometry editing (e.g., inserting speed humps or sunken roads), and generally do not couple such edits with plausible vehicle-road interaction dynamics. Such editing is essential for generating training data under extreme driving scenarios or evaluating system reliability under these road irregularities. Moreover, many optimization-based methods require minutes of per-edit refinement, while existing efficient alternatives mainly focus on appearance-level or object-level manipulation rather than physics-aware road irregularity editing. To address these limitations, we propose RoVES, a Road-and-Vehicle Editing System for physics-aware 3D Gaussian editing in driving scenes. RoVES enables single-image-driven road geometry insertion and couples the edited road profile with a 4-DOF half-car vehicle dynamics model to achieve physics-aware vehicle pose correction in vertical displacement and pitch. RoVES inserts road elements in a one-shot, optimization-free pipeline (1.84s), and the full pipeline (including color transfer and vehicle-dynamics-based pose correction) completes in 6.24s; it edits dynamic vehicles via pose editing and corrects poses frame-by-frame to approximate dynamics-consistent vertical displacement and pitch responses. Experiments on the Waymo dataset show that RoVES provides practical efficiency and competitive visual consistency for physics-aware driving scene generation. △ Less

Submitted 24 May, 2026; originally announced May 2026.

Long Memory in Intrinsically Dynamic Factor Models

Authors: Qin Wen, Clifford M. Hurvich

Abstract: We study the generalized dynamic factor model in a long-memory setting. Unlike most recent work, which assumes a finite-dimensional factor space and short memory, our framework allows the factor space to be infinite-dimensional and the common components to exhibit long memory. We employ the two-sided estimation method of Forni, Hallin, Lippi and Reichlin (2000, Review of Economics and Statistics)… ▽ More We study the generalized dynamic factor model in a long-memory setting. Unlike most recent work, which assumes a finite-dimensional factor space and short memory, our framework allows the factor space to be infinite-dimensional and the common components to exhibit long memory. We employ the two-sided estimation method of Forni, Hallin, Lippi and Reichlin (2000, Review of Economics and Statistics) to recover the common component. The long memory structure of the common component poses a challenge, as it introduces unboundedness/discontinuity in the spectral density. We address this issue by leveraging two key facts: First, the estimated operator is a projection onto the leading eigenspace and thus the eigengap provides an intrinsic scaling that partially mitigates the blow-up. Second, we perform most of our estimation in $L^p$-norm, rather than pointwise. Experimental results are presented to provide evidence supporting the theory, as well as potential improvements to it. △ Less

Submitted 22 May, 2026; originally announced May 2026.

Comments: 92 pages, 11 figures

MSC Class: 62P20 ACM Class: G.3

AutoResearch AI: Towards AI-Powered Research Automation for Scientific Discovery

Authors: Guiyao Tie, Jiawen Shi, Dingjie Song, Yixiao Huang, Ziji Sheng, Xueyang Zhou, Daizong Liu, Pan Zhou, Yongchao Chen, Ran Xu, Lifang He, Qingsong Wen, Manling Li, Cong Lu, Shuai Li, Pengtao Xie, Yixuan Yuan, Rui Meng, Lei Xing, Lichao Sun, Caiming Xiong, Philip S. Yu, Jianfeng Gao

Abstract: Scientific research is being reshaped by AI systems that move beyond isolated assistance toward longer-horizon workflows spanning literature grounding, hypothesis generation, experimentation, validation, reporting, and revision. This shift marks a transition from task-level AI for science to workflow-level research automation. Yet current systems remain fragmented, differing in autonomy, domain sc… ▽ More Scientific research is being reshaped by AI systems that move beyond isolated assistance toward longer-horizon workflows spanning literature grounding, hypothesis generation, experimentation, validation, reporting, and revision. This shift marks a transition from task-level AI for science to workflow-level research automation. Yet current systems remain fragmented, differing in autonomy, domain scope, execution environment, validation mechanism, and human oversight, while still struggling with evidence preservation, reproducibility, weak-direction rejection, provenance tracking, cross-domain robustness, and accountable scientific closure. This survey examines these developments through AutoResearch, defined as the developmental spectrum of AI-powered scientific workflow automation. Within it, Vibe Research denotes the human-steered region of prompt-based assistance and human-verified execution, whereas emerging AI-led systems coordinate larger portions of the discovery loop without achieving robust autonomy. We analyze how research systems redistribute control, evidence, execution, validation, and accountability across workflows and organize the field around five workflow conditions: literature and research grounding; hypothesis formation and planning; experimentation and tool use; feedback, validation, and review; and reporting and knowledge communication. We further synthesize AI scientist systems, mixed-initiative co-research frameworks, benchmarks, domain deployments, and open-source infrastructures. Finally, we propose five evaluation dimensions--novelty, validity, impact, reliability, and provenance--and show that AutoResearch autonomy is domain-conditioned, being more credible in structured, executable, and rapidly verifiable settings but limited in embodied, delayed, heterogeneous, ethical, or institutionally accountable contexts. △ Less

Submitted 21 May, 2026; originally announced May 2026.

Comments: 49 pages, 12 figures, 10 tables

Beyond Extrapolation: Knowledge Utilization Paradigm with Bidirectional Inspiration for Time Series Forecasting

Authors: Liu Chong, Yingjie Zhou, Hao Li, Pengyang Wang, Qingsong Wen, Ce Zhu

Abstract: Time-series forecasting is critical in various scenarios, such as energy, transportation, and public health. However, most existing forecasters rely primarily on one-way inference, \textit{i.e.}, mapping \textbf{history} to \textbf{target}, and overlook the structural information provided by a revised natural chain (``\textbf{history} (model input) -- \textbf{target} (ground-truth output) -- \text… ▽ More Time-series forecasting is critical in various scenarios, such as energy, transportation, and public health. However, most existing forecasters rely primarily on one-way inference, \textit{i.e.}, mapping \textbf{history} to \textbf{target}, and overlook the structural information provided by a revised natural chain (``\textbf{history} (model input) -- \textbf{target} (ground-truth output) -- \textbf{post-target continuation}''). The post-target continuation records how trajectories evolve after the target, which can help stabilize forecasting, but it is not observable at inference time. In this work, we aim to obtain an approximate proxy of the post-target continuation for the current input, providing structural knowledge for bidirectional forecasting. This idea is instantiated as KUP-BI (Knowledge Utilization Paradigm with Bidirectional Inspiration), a new time-series modeling paradigm that distills continuation-style knowledge (as an approximate post-target continuation proxy) from a \emph{train-only} historical library and integrates it into standard forecasting backbones. The input stream and the continuation-proxy stream are fused via a lightweight feature-level gating module. This design does not introduce information beyond what is already contained in the training trajectories; instead, it provides a structured inductive bias that helps backbones exploit typical continuation patterns rather than relying solely on parametric extrapolation. Experimental results on six public datasets show that KUP-BI consistently improves the forecasting performance of state-of-the-art models, with small additional overhead. △ Less

Submitted 18 May, 2026; originally announced May 2026.

Comments: Accepted to ICML 2026. 18 pages, 6 figures

Olivia: Harmonizing Time Series Foundation Models with Power Spectral Density

Authors: Jingru Fei, Kun Yi, Alex Xing Wang, Qingsong Wen, Xiangxiang Zhu, Wei Fan

Abstract: Time series foundation models rely on large-scale pretraining over diverse datasets across domains, yet their heterogeneity in temporal patterns could hinder the effectiveness of training and learning transferable time series representations. Inspired a fundamental concept, normalized power spectral density (PSD) in signal processing, we assume harmonizing datasets via PSDs in the spectral domain… ▽ More Time series foundation models rely on large-scale pretraining over diverse datasets across domains, yet their heterogeneity in temporal patterns could hinder the effectiveness of training and learning transferable time series representations. Inspired a fundamental concept, normalized power spectral density (PSD) in signal processing, we assume harmonizing datasets via PSDs in the spectral domain could reduce mismatches and enhance pretraining. We then go beyond the direct intractable minimization optimization and innovatively reformulate it as a principled harmonization approach. Specifically, we propose Harmonizer, a module that reshapes spectral structures and implicitly harmonizing PSDs across datasets, which theoretically corresponds to a shared reparameterization of second-order temporal correlations. Our theoretical analysis further reveals token interactions with Harmonizer can be efficiently mediated by a compact set of resonators, motivating a HarmonicAttention design that performs self-attention in a low-dimensional interaction space. Then, we propose Olivia, a novel time series foundation model built upon these harmonization mechanisms. Extensive experiments on two large-scale benchmarks (TSLib and GIFT-Eval) and extra 6 datasets from GluonTS, demonstrate Olivia consistently achieves state-of-the-art performance under zero-shot, few-shot, and full-shot forecasting scenarios. Our code is available at https://github.com/TSTS13/Olivia. △ Less

Submitted 18 May, 2026; v1 submitted 17 May, 2026; originally announced May 2026.

Comments: Accepted by ICML 2026

CHI-Bench: Can AI Agents Automate End-to-End, Long-Horizon, Policy-Rich Healthcare Workflows?

Authors: Haolin Chen, Deon Metelski, Leon Qi, Tao Xia, Joonyul Lee, Steve Brown, Kevin Riley, Frank Wang, T. Y. Alvin Liu, Hank Capps MD, Zeyu Tang, Xiangchen Song, Lingjing Kong, Fan Feng, Tianyi Zeng, Zhiwei Liu, Zixian Ma, Hang Jiang, Fangli Geng, Yuan Yuan, Chenyu You, Qingsong Wen, Hua Wei, Yanjie Fu, Yue Zhao , et al. (8 additional authors not shown)

Abstract: End-to-end automation of realistic healthcare operations stresses three capabilities underrepresented in current benchmarks: policy density, decisions must be grounded in a large library of medical, insurance, and operational rules; Multi-role composition: a single task requires the agent to play multiple roles with handoffs; and multilateral interaction: intermediate workflow steps are multi-turn… ▽ More End-to-end automation of realistic healthcare operations stresses three capabilities underrepresented in current benchmarks: policy density, decisions must be grounded in a large library of medical, insurance, and operational rules; Multi-role composition: a single task requires the agent to play multiple roles with handoffs; and multilateral interaction: intermediate workflow steps are multi-turn dialogs, such as peer-to-peer review and patient outreach. We introduce $χ$-Bench, a benchmark of long-horizon healthcare workflows across three domains: provider prior authorization, payer utilization management, and care management. Each task hands the agent a clinical case in a high-fidelity simulator of 20 healthcare apps exposed via 87 MCP tools, which it must drive to a terminal status through tool calls and writing the role's artifacts, guided by a 1,290+ document managed-care operations handbook skill. Across 30 agent harness/models configurations, the best agent resolves only 28.0% of tasks, no agent clears 20% on strict pass^3, and executing all tasks in a single session slumps the performance to 3.8%. These results raise the hypothesis that similar gaps are likely to surface in other policy-dense, role-composed, irreversible enterprise domains. △ Less

Submitted 19 May, 2026; v1 submitted 15 May, 2026; originally announced May 2026.

Comments: Website: https://actava.ai/benchmarks Code: https://github.com/actava-ai/chi-bench Dataset: https://huggingface.co/datasets/actava/chi-bench

Robust Sequential Experimental Design for A/B Testing

Authors: Qianglin Wen, Xiangkun Wu, Chengchun Shi, Ting Li, Niansheng Tang, Yingying Zhang, Hongtu Zhu

Abstract: Experimental design has emerged as a powerful approach for improving the sample efficiency of A/B testing, yet existing designs rely critically on correctly specified models. We study robust sequential experimental design under model misspecification and develop a unified framework that covers both contextual bandit and dynamic settings. Theoretically, we prove that our design bounds the worst-cas… ▽ More Experimental design has emerged as a powerful approach for improving the sample efficiency of A/B testing, yet existing designs rely critically on correctly specified models. We study robust sequential experimental design under model misspecification and develop a unified framework that covers both contextual bandit and dynamic settings. Theoretically, we prove that our design bounds the worst-case mean squared error of the estimated treatment effect. Empirically, we demonstrate the effectiveness of the proposed approach using synthetic and real-world datasets from a leading technology company. △ Less

Submitted 12 May, 2026; originally announced May 2026.

UniPath: Adaptive Coordination of Understanding and Generation for Unified Multimodal Reasoning

Authors: Hayes Bai, Yinyi Luo, Wenwen Wang, Qingsong Wen, Jindong Wang

Abstract: Unified multimodal models (UMMs) aim to integrate understanding and generation within a single architecture. However, it remains underexplored how to effectively coordinate these two capabilities for more effective and efficient reasoning. Existing coordination approaches either perform coupling during training, without explicit inference-time coordination, or impose a fixed coordination pattern f… ▽ More Unified multimodal models (UMMs) aim to integrate understanding and generation within a single architecture. However, it remains underexplored how to effectively coordinate these two capabilities for more effective and efficient reasoning. Existing coordination approaches either perform coupling during training, without explicit inference-time coordination, or impose a fixed coordination pattern for all inputs. In this work, we show that multimodal tasks exhibit substantial coordination-path diversity: different inputs favor different coordination paths. This suggests that exploiting such diversity is key to improving performance. We propose UniPath, a framework for adaptively modeling and exploiting coordination-path diversity. Instead of enforcing a single coordination pattern, we represent task solving as the selection and execution of a path, ranging from direct answering to textual inference, visual-thought construction, and hypothesis-based exploration. We construct role-aligned trajectories to train a path-conditioned executor and introduce a lightweight planner mechanism to enable input-dependent path selection. Experiments show that leveraging coordination-path diversity improves performance over fixed coordination strategies while providing interpretable intermediate behaviors. The code is available at:https://github.com/AIFrontierLab/TorchUMM/tree/main/src/umm/post_training/unipath. △ Less

Submitted 11 May, 2026; originally announced May 2026.

Remember the Decision, Not the Description: A Rate-Distortion Framework for Agent Memory

Authors: Mingxi Zou, Zhihan Guo, Langzhang Liang, Zhuo Wang, Qifan Wang, Qingsong Wen, Irwin King, Lizhen Qu, Zenglin Xu

Abstract: Long-horizon language agents must operate under limited runtime memory, yet existing memory mechanisms often organize experience around descriptive criteria such as relevance, salience, or summary quality. For an agent, however, memory is valuable not because it faithfully describes the past, but because it preserves the distinctions between histories that must remain separated under a fixed budge… ▽ More Long-horizon language agents must operate under limited runtime memory, yet existing memory mechanisms often organize experience around descriptive criteria such as relevance, salience, or summary quality. For an agent, however, memory is valuable not because it faithfully describes the past, but because it preserves the distinctions between histories that must remain separated under a fixed budget to support good decisions. We cast this as a decision-centric rate-distortion problem, measuring memory quality by the loss in achievable decision quality induced by compression. This yields an exact forgetting boundary for what can be safely forgotten, and a memory-distortion frontier characterizing the optimal tradeoff between memory budget and decision quality. Motivated by this decision-centric view of memory, we propose DeMem, an online memory learner that refines its partition only when data certify that a shared state would induce decision conflict, and prove near-minimax regret guarantees. On both controlled synthetic diagnostics and long-horizon conversational benchmarks, DeMem yields consistent gains under the same runtime budget, supporting the principle that memory should preserve the distinctions that matter for decisions, not descriptions. △ Less

Submitted 11 May, 2026; originally announced May 2026.

PnP-Corrector: A Universal Correction Framework for Coupled Spatiotemporal Forecasting

Authors: Hao Wu, Fan Xu, Yuxu Lu, Penghao Zhao, Fan Zhang, Hao Jia, Yuxuan Liang, Ruijian Gou, Qingsong Wen, Xian Wu, Xiaomeng Huang, Yuan Gao

Abstract: Coupled spatiotemporal forecasting is important for predicting the future evolution of multiple interacting dynamical systems, such as in climate models. However, existing methods are severely constrained by the persistent bottleneck of compounding errors. In coupled systems, errors from each subsystem simulator propagate and amplify one another, a phenomenon we term Reciprocal Error Amplification… ▽ More Coupled spatiotemporal forecasting is important for predicting the future evolution of multiple interacting dynamical systems, such as in climate models. However, existing methods are severely constrained by the persistent bottleneck of compounding errors. In coupled systems, errors from each subsystem simulator propagate and amplify one another, a phenomenon we term Reciprocal Error Amplification, leading to a rapid collapse of long-range predictions. To address this challenge, we propose a universal framework called PnP-Corrector (Plug-and-Play Corrector). The core idea of our framework is to decouple the physical simulation from the error correction process: it freezes pre-trained physics simulation engines and exclusively trains a correction agent to proactively counteract the systematic biases emerging from the coupled system. Furthermore, we design an efficient predictive model architecture, DSLCast, to serve as the backbone of this framework. Extensive experiments demonstrate that our method significantly enhances the long-term stability and accuracy of coupled forecasting systems. For instance, in the challenging task of a 300-day global ocean-atmosphere coupled forecast, our PnP-Corrector framework reduces the prediction error of the baseline model by 28% and surpasses state-of-the-art models on several key metrics. △ Less

Submitted 2 June, 2026; v1 submitted 9 May, 2026; originally announced May 2026.

Perceive, Route and Modulate: Dynamic Pattern Recalibration for Time Series Forecasting

Authors: Siru Zhong, Zhao Meng, Haohuan Fu, Haoyang Li, Qingsong Wen, Yuxuan Liang

Abstract: Local temporal patterns in real-world time series continuously shift, rendering globally shared transformations suboptimal. Current deep forecasting models, despite their scale and complexity, rely on fixed weight matrices applied uniformly to all temporal tokens. This creates a static pattern response: models settle into a compromised average, unable to adapt to changing local dynamics. We introd… ▽ More Local temporal patterns in real-world time series continuously shift, rendering globally shared transformations suboptimal. Current deep forecasting models, despite their scale and complexity, rely on fixed weight matrices applied uniformly to all temporal tokens. This creates a static pattern response: models settle into a compromised average, unable to adapt to changing local dynamics. We introduce Dynamic Pattern Recalibration (DPR), a backbone-agnostic mechanism that resolves this via token-level recalibration. Through a lightweight "Perceive-Route-Modulate" pipeline, DPR computes a soft-routing distribution over a learned basis of adaptive response patterns, generating a time-aware modulation vector that recalibrates hidden states via a residual Hadamard product. As a backbone-agnostic adapter, DPR enhances forecasting across diverse architectures with minimal overhead, confirming it addresses a general bottleneck. As a minimalist standalone model, DPRNet achieves competitive performance across 12 benchmarks, validating dynamic recalibration against macroscopic parameter scaling. △ Less

Submitted 7 May, 2026; originally announced May 2026.

Comments: 22 pages, 6 figures. Preprint

MIRL: Mutual Information-Guided Reinforcement Learning for Vision-Language Models

Authors: Yin Zhang, Jiaxuan Zhao, Zonghan Wu, Zengxiang Li, Junfeng Fang, Kun Wang, Qingsong Wen, Yilei Shao

Abstract: Vision-Language Models (VLMs) frequently suffer from visual perception errors and hallucinations that compromise answer accuracy in complex reasoning tasks. Reinforcement Learning with Verifiable Rewards (RLVR) offers a promising solution by optimizing policies using answer correctness signals. Despite their effectiveness, prevailing RLVR methods face two critical limitations. First, much of the s… ▽ More Vision-Language Models (VLMs) frequently suffer from visual perception errors and hallucinations that compromise answer accuracy in complex reasoning tasks. Reinforcement Learning with Verifiable Rewards (RLVR) offers a promising solution by optimizing policies using answer correctness signals. Despite their effectiveness, prevailing RLVR methods face two critical limitations. First, much of the sampling budget is wasted on trajectories doomed to fail due to early visual description errors. Second, sparse rewards cannot distinguish whether failures stem from visual perception or reasoning stages. We introduce MIRL, a decoupled framework that addresses both limitations by leveraging mutual information (MI) between generated descriptions and visual inputs as a cheap pre-screening signal. This enables intelligent budget allocation toward high-potential trajectories via forking, while decoupled training provides independent MI-based rewards for visual perception optimization, resolving reward blindness. Experiments on six vision-language reasoning benchmarks demonstrate that MIRL achieves 70.22% average accuracy and successfully surpasses the performance of sampling 16 complete trajectories using only 10 pre-samples with top-6 selection (25% fewer complete trajectories). Our code is available at: https://anonymous.4open.science/r/mirl-main/. △ Less

Submitted 2 May, 2026; originally announced May 2026.

Benchmarking virtual cell models for in-the-wild perturbation response

Authors: Xinjie Mao, Songming Zhang, Qianhong Wen, Xiangyu Wen, Kedu Jin, Hao Wu, Shuizhou Chen, Yuqiang Li, Lei Bai, Qi Liu, Ning Ding, Siqi Sun, Zhangyang Gao

Abstract: Virtual cell (VC) models aim to predict cellular responses to any perturbations in silico and have emerged as a promising approach for drug discovery and precision medicine. Yet, a clear gap still remains: while models routinely reported impressive results on standard benchmarks, it is unclear whether their predictions are truly meaningful in practice. This is mainly due to limitations in current… ▽ More Virtual cell (VC) models aim to predict cellular responses to any perturbations in silico and have emerged as a promising approach for drug discovery and precision medicine. Yet, a clear gap still remains: while models routinely reported impressive results on standard benchmarks, it is unclear whether their predictions are truly meaningful in practice. This is mainly due to limitations in current evaluation setups, which are often overly simplified or inconsistent, and do not reflect the complexity and variability of real biological systems. Here, we introduce a standardized and modular benchmarking framework for virtual cell prediction. Our framework evaluates diverse models under in-the-wild challenging scenarios, including unseen cell contexts, unseen perturbations, and cross-dataset generalization, which better reflect practical applications. Our analysis shows that model performance is highly context-dependent and shaped by task design and evaluation criteria. In commonly used setups, performance is often overestimated, and naive dataset aggregation can even reduce performance. When evaluated under more strict conditions, model performance drops markedly, indicating limited robustness to shifts across cellular contexts. In unseen perturbation settings, models including simple linear approaches capture global transcriptional trends but fail to recover fine-grained perturbation-specific effects. In addition, different evaluation metrics focus on different biological properties, leading to substantially different model rankings. Together, our framework provides a more reliable and biologically grounded evaluation, offering clearer guidance for applying virtual cell models in real scenarios. △ Less

Submitted 30 April, 2026; originally announced April 2026.

Designing and Evaluating Next-Generation Learning Interfaces: Linking AI, HCI, and the Learning Sciences

Authors: Meng Xia, Yan Chen, Qiao Jin, Yang Shi, Paul Denny, Tiffany Barnes, Qingsong Wen, Vincent Aleven

Abstract: This workshop addresses this gap by bringing together researchers and practitioners from AI, HCI, and the learning sciences to explore how interactive systems can better support learning. We focus on the design and evaluation of human-AI collaborative learning interfaces that are technically robust, human-centered, and pedagogically grounded. By fostering interdisciplinary dialogue, the workshop a… ▽ More This workshop addresses this gap by bringing together researchers and practitioners from AI, HCI, and the learning sciences to explore how interactive systems can better support learning. We focus on the design and evaluation of human-AI collaborative learning interfaces that are technically robust, human-centered, and pedagogically grounded. By fostering interdisciplinary dialogue, the workshop aims to identify shared challenges, design principles, and research directions for next-generation learning technologies. △ Less

Submitted 28 April, 2026; originally announced April 2026.

End-to-End Learning for Partially-Observed Time Series with PyPOTS

Authors: Wenjie Du, Yiyuan Yang, Tianxiang Zhan, Qingsong Wen

Abstract: Partially-observed time series (POTS) is ubiquitous in real-world applications, yet most existing toolchains separate missing-value handling from downstream learning, which limits reproducibility and overall performance. This tutorial introduces PyPOTS, an open-source Python ecosystem for end-to-end data mining and machine learning on POTS. We present practical workflows spanning missingness simul… ▽ More Partially-observed time series (POTS) is ubiquitous in real-world applications, yet most existing toolchains separate missing-value handling from downstream learning, which limits reproducibility and overall performance. This tutorial introduces PyPOTS, an open-source Python ecosystem for end-to-end data mining and machine learning on POTS. We present practical workflows spanning missingness simulation, data preprocessing, model training, and evaluation across core tasks, including imputation, forecasting, classification, clustering, and anomaly detection. The tutorial consists of two parts: Part I emphasizes hands-on application for practitioners through unified APIs and benchmark-oriented experiments. Part II targets developers and researchers, focusing on extending PyPOTS with custom models, domain-specific constraints, and contribution-ready engineering practices. Participants will gain both conceptual understanding and implementation experience for building robust, transparent, and reusable POTS pipelines in research and production settings. PyPOTS is publicly available at https://github.com/WenjieDu/PyPOTS △ Less

Submitted 27 April, 2026; originally announced April 2026.

Comments: Accepted by KDD 2026

Bayesian Active Learning with Gaussian Processes Guided by LLM Relevance Scoring for Dense Passage Retrieval

Authors: Junyoung Kim, Anton Korikov, Jiazhou Liang, Justin Cui, Yifan Simon Liu, Qianfeng Wen, Mark Zhao, Scott Sanner

Abstract: While Large Language Models (LLMs) exhibit exceptional zero-shot relevance modeling, their high computational cost necessitates framing passage retrieval as a budget-constrained global optimization problem. Existing approaches passively rely on first-stage dense retrievers, which leads to two limitations: (1) failing to retrieve relevant passages in semantically distinct clusters, and (2) failing… ▽ More While Large Language Models (LLMs) exhibit exceptional zero-shot relevance modeling, their high computational cost necessitates framing passage retrieval as a budget-constrained global optimization problem. Existing approaches passively rely on first-stage dense retrievers, which leads to two limitations: (1) failing to retrieve relevant passages in semantically distinct clusters, and (2) failing to propagate relevance signals to the broader corpus. To address these limitations, we propose Bayesian Active Learning with Gaussian Processes guided by LLM relevance scoring (BAGEL), a novel framework that propagates sparse LLM relevance signals across the embedding space to guide global exploration. BAGEL models the multimodal relevance distribution across the entire embedding space with a query-specific Gaussian Process (GP) based on LLM relevance scores. Subsequently, it iteratively selects passages for scoring by strategically balancing the exploitation of high-confidence regions with the exploration of uncertain areas. Extensive experiments across four benchmark datasets and two LLM backbones demonstrate that BAGEL effectively explores and captures complex relevance distributions and outperforms LLM reranking methods under the same LLM budget on all four datasets. △ Less

Submitted 20 April, 2026; originally announced April 2026.

Comments: ACL 2026 Findings

Journal ref: Findings of the Association for Computational Linguistics: ACL 2026

Effects of Soret diffusion on the intrinsic instability of premixed hydrogen/air flames

Authors: Qizhe Wen, Yan Wang, Linlin Yang, Youhi Morii, Thorsten Zirwes, Shengkai Wang, Zheng Chen

Abstract: Hydrogen flames exhibit multiple intrinsic instabilities. The low molar masses of H and H2 lead to significant Soret diffusion near the flame front; however, its influence on hydrogen flame instabilities remains to be quantified. This study investigates the effect of Soret diffusion on instability evolution dynamics via one-dimensional counterflow analysis and two-dimensional, high-fidelity direct… ▽ More Hydrogen flames exhibit multiple intrinsic instabilities. The low molar masses of H and H2 lead to significant Soret diffusion near the flame front; however, its influence on hydrogen flame instabilities remains to be quantified. This study investigates the effect of Soret diffusion on instability evolution dynamics via one-dimensional counterflow analysis and two-dimensional, high-fidelity direct numerical simulations covering both the linear growth regime and the fully developed nonlinear regime over a wide range of equivalence ratios (phi). In the linear regime, Soret diffusion increases the perturbation growth rate at phi < 1.7, especially under lean conditions, but reduces the growth rate at phi > 1.7. A similar sensitivity reversal is observed in the Markstein length near the peak equivalence ratio of unstretched laminar flame speed. In the nonlinear regime, Soret diffusion accelerates the formation of small-scale wrinkles in lean hydrogen flames and reduces the characteristic size of large-scale finger structure by one-third. An interesting observation is that, although Soret diffusion promotes preferential diffusion and increases the local flame displacement speed, the global fuel consumption rate decreases due to a reduction in the overall flame surface area. In addition, curvature-based flame segment analysis reveals a synergistic effect between Soret diffusion and Fickian diffusion that enhances/reduces the local equivalence ratio in positively/negatively curved regions of the flame front. The probability distributions of the Karlovitz number and the density-weighted displacement speed are also analyzed; results suggest that, for lean hydrogen flames, Soret diffusion broadens the distributions for both parameters, particularly on the positive side. These findings promise to advance the fundamental understanding of hydrogen flame dynamics under complex differential transport. △ Less

Submitted 5 July, 2026; v1 submitted 9 April, 2026; originally announced April 2026.

ThinkTwice: Jointly Optimizing Large Language Models for Reasoning and Self-Refinement

Authors: Difan Jiao, Qianfeng Wen, Blair Yang, Zhenwei Tang, Ashton Anderson

Abstract: We introduce ThinkTwice, a simple two-phase framework that jointly optimizes LLMs to solve reasoning problems and refine the answers, based on Group Relative Policy Optimization (GRPO). In each pair of training steps, ThinkTwice first optimizes the model on solving reasoning problems, then optimizes it on refining its own solutions to the same problems, using the same binary correctness reward in… ▽ More We introduce ThinkTwice, a simple two-phase framework that jointly optimizes LLMs to solve reasoning problems and refine the answers, based on Group Relative Policy Optimization (GRPO). In each pair of training steps, ThinkTwice first optimizes the model on solving reasoning problems, then optimizes it on refining its own solutions to the same problems, using the same binary correctness reward in both phases without correctness signals or critique annotations. Across five mathematical reasoning benchmarks and two model families including Qwen3-4B and Olmo3-7B, ThinkTwice substantially improves both reasoning and refinement performance over competitive online policy optimization baselines. Specifically, on Qwen3-4B, ThinkTwice outperforms GRPO on AIME by 5 percentage points before refinement and by 11.5 points after one self-refinement step, measured by pass@4. Analysis of the training dynamics of ThinkTwice reveals an implicit rectify-then-fortify curriculum: refinement predominantly corrects errors early in training and naturally shifts toward preserving already-correct solutions as the model improves, yielding a more rectified reward signal. Our work establishes joint training of reasoning and self-refinement as a principled and effective methodology for RLVR. △ Less

Submitted 6 April, 2026; v1 submitted 2 April, 2026; originally announced April 2026.

Comments: 27 pages,7 figures,5 tables

Bridging the Geometry Mismatch: Frequency-Aware Anisotropic Serialization for Thin-Structure SSMs

Authors: Jin Bai, Huiyao Zhang, Qi Wen, Ningyang Li, Shengyang Li, Atta ur Rahman, Xiaolin Tian

Abstract: The segmentation of thin linear structures is inherently topology allowbreak-critical, where minor local errors can sever long-range connectivity. While recent State-Space Models (SSMs) offer efficient long-range modeling, their isotropic serialization (e.g., raster scanning) creates a geometry mismatch for anisotropic targets, causing state propagation across rather than along the structure traje… ▽ More The segmentation of thin linear structures is inherently topology allowbreak-critical, where minor local errors can sever long-range connectivity. While recent State-Space Models (SSMs) offer efficient long-range modeling, their isotropic serialization (e.g., raster scanning) creates a geometry mismatch for anisotropic targets, causing state propagation across rather than along the structure trajectories. To address this, we propose FGOS-Net, a framework based on frequency allowbreak-geometric disentanglement. We first decompose features into a stable topology carrier and directional high-frequency bands, leveraging the latter to explicitly correct spatial misalignments induced by downsampling. Building on this calibrated topology, we introduce frequency-aligned scanning that elevates serialization to a geometry-conditioned decision, preserving direction-consistent traces. Coupled with an active probing strategy to selectively inject high-frequency details and suppress texture ambiguity, FGOS-Net consistently outperforms strong baselines across four challenging benchmarks. Notably, it achieves 91.3% mIoU and 97.1% clDice on DeepCrack while running at 80 FPS with only 7.87 GFLOPs. △ Less

Submitted 30 March, 2026; originally announced March 2026.

Effects of gravity on lean hydrogen/air flame instability: From linear scaling law to nonlinear morphology evolution

Authors: Qizhe Wen, Yan Wang, Linlin Yang, Yiqing Wang, Thorsten Zirwes, Shengkai Wang, Zheng Chen

Abstract: The instability characteristics of lean hydrogen/air flames have attracted considerable research attention, yet the effect of gravity remains insufficiently understood. In this study, time-resolved two-dimensional simulations with detailed chemistry and transport are conducted to investigate the influence of gravity-induced Rayleigh-Taylor (RT) instability on the linear growth rate of disturbances… ▽ More The instability characteristics of lean hydrogen/air flames have attracted considerable research attention, yet the effect of gravity remains insufficiently understood. In this study, time-resolved two-dimensional simulations with detailed chemistry and transport are conducted to investigate the influence of gravity-induced Rayleigh-Taylor (RT) instability on the linear growth rate of disturbances and nonlinear morphology evolution of cellular flame fronts at different length scales. In the linear regime, a parametric study is performed across various equivalence ratios, initial temperatures and pressures; in each case, the dispersion relation is calculated for various gravity levels. The influence of gravity is most pronounced under ultra-lean, low-temperature, and high-pressure conditions, and a universal scaling law between gravity sensitivity and the Froude number is established. In the nonlinear regime, gravity has opposite effects on the large-scale and small-scale structures of lean hydrogen flames. On the one hand, gravity inhibits the splitting of small-scale cellular structures through a baroclinic torque mechanism; on the other hand, it promotes the development of large-scale finger-like structures, thereby increasing the total surface area and the global consumption speed of the flame. The effects of gravity on the probability distributions of cell size, displacement speed, Karlovitz number, and local curvature are also analyzed. The results and findings of the present study should advance the fundamental understanding of hydrogen flame dynamics under varying gravity conditions and provide insight for relevant applications, including fire safety and space propulsion. △ Less

Submitted 30 March, 2026; originally announced March 2026.

Can MLLMs Read Students' Minds? Unpacking Multimodal Error Analysis in Handwritten Math

Authors: Dingjie Song, Tianlong Xu, Yi-Fan Zhang, Hang Li, Zhiling Yan, Xing Fan, Haoyang Li, Lichao Sun, Qingsong Wen

Abstract: Assessing student handwritten scratchwork is crucial for personalized educational feedback but presents unique challenges due to diverse handwriting, complex layouts, and varied problem-solving approaches. Existing educational NLP primarily focuses on textual responses and neglects the complexity and multimodality inherent in authentic handwritten scratchwork. Current multimodal large language mod… ▽ More Assessing student handwritten scratchwork is crucial for personalized educational feedback but presents unique challenges due to diverse handwriting, complex layouts, and varied problem-solving approaches. Existing educational NLP primarily focuses on textual responses and neglects the complexity and multimodality inherent in authentic handwritten scratchwork. Current multimodal large language models (MLLMs) excel at visual reasoning but typically adopt an "examinee perspective", prioritizing generating correct answers rather than diagnosing student errors. To bridge these gaps, we introduce ScratchMath, a novel benchmark specifically designed for explaining and classifying errors in authentic handwritten mathematics scratchwork. Our dataset comprises 1,720 mathematics samples from Chinese primary and middle school students, supporting two key tasks: Error Cause Explanation (ECE) and Error Cause Classification (ECC), with seven defined error types. The dataset is meticulously annotated through rigorous human-machine collaborative approaches involving multiple stages of expert labeling, review, and verification. We systematically evaluate 16 leading MLLMs on ScratchMath, revealing significant performance gaps relative to human experts, especially in visual recognition and logical reasoning. Proprietary models notably outperform open-source models, with large reasoning models showing strong potential for error explanation. All evaluation data and frameworks are publicly available to facilitate further research. △ Less

Submitted 25 March, 2026; originally announced March 2026.

Comments: Accepted by the 27th International Conference on Artificial Intelligence in Education (AIED'26)

ACM Class: I.2.7; K.3.1

SafeSeek: Universal Attribution of Safety Circuits in Language Models

Authors: Miao Yu, Siyuan Fu, Moayad Aloqaily, Zhenhong Zhou, Safa Otoum, Xing fan, Kun Wang, Yufei Guo, Qingsong Wen

Abstract: Mechanistic interpretability reveals that safety-critical behaviors (e.g., alignment, jailbreak, backdoor) in Large Language Models (LLMs) are grounded in specialized functional components. However, existing safety attribution methods struggle with generalization and reliability due to their reliance on heuristic, domain-specific metrics and search algorithms. To address this, we propose \ourmetho… ▽ More Mechanistic interpretability reveals that safety-critical behaviors (e.g., alignment, jailbreak, backdoor) in Large Language Models (LLMs) are grounded in specialized functional components. However, existing safety attribution methods struggle with generalization and reliability due to their reliance on heuristic, domain-specific metrics and search algorithms. To address this, we propose \ourmethod, a unified safety interpretability framework that identifies functionally complete safety circuits in LLMs via optimization. Unlike methods focusing on isolated heads or neurons, \ourmethod introduces differentiable binary masks to extract multi-granular circuits through gradient descent on safety datasets, while integrates Safety Circuit Tuning to utilize these sparse circuits for efficient safety fine-tuning. We validate \ourmethod in two key scenarios in LLM safety: \textbf{(1) backdoor attacks}, identifying a backdoor circuit with 0.42\% sparsity, whose ablation eradicates the Attack Success Rate (ASR) from 100\% $\to$ 0.4\% while retaining over 99\% general utility; \textbf{(2) safety alignment}, localizing an alignment circuit with 3.03\% heads and 0.79\% neurons, whose removal spikes ASR from 0.8\% $\to$ 96.9\%, whereas excluding this circuit during helpfulness fine-tuning maintains 96.5\% safety retention. △ Less

Submitted 24 March, 2026; originally announced March 2026.

Grounded Chess Reasoning in Language Models via Master Distillation

Authors: Zhenwei Tang, Qianfeng Wen, Seth Grief-Albert, Yahya Elgabra, Blair Yang, Honghua Dong, Ashton Anderson

Abstract: Language models often lack grounded reasoning capabilities in specialized domains where training data is scarce but bespoke systems excel. We introduce a general framework for distilling expert system reasoning into natural language chain-of-thought explanations, enabling compact models to acquire domain expertise and the ability to generate faithful, grounded explanations. Rather than distilling… ▽ More Language models often lack grounded reasoning capabilities in specialized domains where training data is scarce but bespoke systems excel. We introduce a general framework for distilling expert system reasoning into natural language chain-of-thought explanations, enabling compact models to acquire domain expertise and the ability to generate faithful, grounded explanations. Rather than distilling only final outputs, we capture the full reasoning process, transforming opaque expert computations into transparent, step-by-step explanations. We demonstrate this approach in chess, a canonical reasoning domain where language models continue to underperform. Our 4B parameter model, C1, advances from a near-zero baseline to 48.1\% accuracy, outperforming all open-source models and most frontier proprietary systems. Notably, C1 surpasses its distillation teacher and generates solutions in two orders of magnitude fewer tokens than baselines. Unlike prior neural chess approaches that predict only best moves, C1 generates explainable solutions revealing strategic reasoning. Our pipeline combines supervised fine-tuning and reinforcement learning with theme-balanced data sampling for comprehensive tactical coverage. Master Distillation demonstrates how to inject expert-level knowledge into compact models for under-optimized domains, offering a recipe for unlocking RLVR where LLMs lack sufficient base capabilities. △ Less

Submitted 23 June, 2026; v1 submitted 20 March, 2026; originally announced March 2026.

Deep Autocorrelation Modeling for Time-Series Forecasting: Progress and Prospects

Authors: Hao Wang, Licheng Pan, Qingsong Wen, Jialin Yu, Zhichao Chen, Chunyuan Zheng, Xiaoxi Li, Zhixuan Chu, Chao Xu, Mingming Gong, Haoxuan Li, Yuan Lu, Zhouchen Lin, Philip Torr, Yan Liu

Abstract: Autocorrelation is a defining characteristic of time-series data, where each observation is statistically dependent on its predecessors. In the context of deep time-series forecasting, autocorrelation arises in both the input history and the label sequences, presenting two central research challenges: (1) designing neural architectures that model autocorrelation in history sequences, and (2) devis… ▽ More Autocorrelation is a defining characteristic of time-series data, where each observation is statistically dependent on its predecessors. In the context of deep time-series forecasting, autocorrelation arises in both the input history and the label sequences, presenting two central research challenges: (1) designing neural architectures that model autocorrelation in history sequences, and (2) devising learning objectives that model autocorrelation in label sequences. Recent studies have made strides in tackling these challenges, but a systematic survey examining both aspects remains lacking. To bridge this gap, this paper provides a comprehensive review of deep time-series forecasting from the perspective of autocorrelation modeling. In contrast to existing surveys, this work makes two distinctive contributions. First, it proposes a novel taxonomy that encompasses recent literature on both model architectures and learning objectives -- whereas prior surveys neglect or inadequately discuss the latter aspect. Second, it offers a thorough analysis of the motivations, insights, and progression of the surveyed literature from a unified, autocorrelation-centric perspective, providing a holistic overview of the evolution of deep time-series forecasting. The full list of papers and resources is available at https://github.com/Master-PLC/Awesome-TSF-Papers. △ Less

Submitted 20 March, 2026; originally announced March 2026.

Orthogonal Subspace Clustering: Enhancing High-Dimensional Data Analysis through Adaptive Dimensionality Reduction and Efficient Clustering

Authors: Qing-Yuan Wen, Da-Qing Zhang

Abstract: This paper presents Orthogonal Subspace Clustering (OSC), an innovative method for high-dimensional data clustering. We first establish a theoretical theorem proving that high-dimensional data can be decomposed into orthogonal subspaces in a statistical sense, whose form exactly matches the paradigm of Q-type factor analysis. This theorem lays a solid mathematical foundation for dimensionality red… ▽ More This paper presents Orthogonal Subspace Clustering (OSC), an innovative method for high-dimensional data clustering. We first establish a theoretical theorem proving that high-dimensional data can be decomposed into orthogonal subspaces in a statistical sense, whose form exactly matches the paradigm of Q-type factor analysis. This theorem lays a solid mathematical foundation for dimensionality reduction via matrix decomposition and factor analysis. Based on this theorem, we propose the OSC framework to address the "curse of dimensionality" -- a critical challenge that degrades clustering effectiveness due to sample sparsity and ineffective distance metrics. OSC integrates orthogonal subspace construction with classical clustering techniques, introducing a data-driven mechanism to select the subspace dimension based on cumulative variance contribution. This avoids manual selection biases while maximizing the retention of discriminative information. By projecting high-dimensional data into an uncorrelated, low-dimensional orthogonal subspace, OSC significantly improves clustering efficiency, robustness, and accuracy. Extensive experiments on various benchmark datasets demonstrate the effectiveness of OSC, with thorough analysis of evaluation metrics including Cluster Accuracy (ACC), Normalized Mutual Information (NMI), and Adjusted Rand Index (ARI) highlighting its advantages over existing methods. △ Less

Submitted 15 March, 2026; originally announced March 2026.