Showing 1–50 of 352 results for author: Fan, M

Trust Is Not Enough: Influence Calibration for On-Policy Self-Distillation in Agentic RL

Authors: Qizhen Lan, Xi Xiao, Xiangchen Guan, Mengchen Fan, Moule Lin, Jung Im Choi, Lijing Zhu

Abstract: On-policy self-distillation (OPSD) gives language agents dense token-level supervision from a privileged self-teacher on the policy's own trajectories. Existing methods allocate this supervision mainly by teacher trust, but trust does not reveal whether emphasizing a token supports the current policy objective. We call this the trust-utility mismatch and introduce Influence Calibration for Self-Di… ▽ More On-policy self-distillation (OPSD) gives language agents dense token-level supervision from a privileged self-teacher on the policy's own trajectories. Existing methods allocate this supervision mainly by teacher trust, but trust does not reveal whether emphasizing a token supports the current policy objective. We call this the trust-utility mismatch and introduce Influence Calibration for Self-Distillation (ICSD). For each supervised token, ICSD measures the first-order response of its importance-weighted RL surrogate contribution to a teacher-directed output perturbation. Batch-adaptive calibration converts this non-stationary signal into a bounded allocation weight while preserving the original auxiliary-loss mass within each action turn. These detached weights affect only the distillation loss and require no additional model pass. Across ALFWorld, WebShop, and Search-QA, ICSD improves all matched aggregate metrics over trust-only allocation under Group Relative Policy Optimization (GRPO) and Group-in-Group Policy Optimization (GiGPO), across two model families spanning 1.5B to 7B. At 7B, it reaches 96.1% ALFWorld success and a WebShop score of 93.1. Frozen-batch analyses show that ICSD reduces teacher-supported mass assigned to objective-opposed tokens from 60.1% to 37.8% and raises cosine compatibility with the RL gradient by 0.192. A companion repository is avail- able at https://github.com/lanqz7766/Influence-Calibration-for-On-Policy-Self-Distillation-in-Agentic-RL. △ Less

Submitted 14 August, 2026; originally announced August 2026.

Control of hybrid wind-wave energy systems using reinforcement learning

Authors: Zechuan Lin, Kemeng Chen, Maosen Fan, Xiaofan Li, Xi Xiao, John V. Ringwood

Abstract: Integrating wave energy converters (WECs) with floating offshore wind turbines (FOWTs), to form hybrid wind-wave energy (HWWE) systems, is a promising approach to achieve further cost reduction for offshore renewable energy. In such systems, the control of the integrated WECs plays an important role, with the potential to generate additional wave energy while simultaneously suppressing floating pl… ▽ More Integrating wave energy converters (WECs) with floating offshore wind turbines (FOWTs), to form hybrid wind-wave energy (HWWE) systems, is a promising approach to achieve further cost reduction for offshore renewable energy. In such systems, the control of the integrated WECs plays an important role, with the potential to generate additional wave energy while simultaneously suppressing floating platform motion. However, HWWE systems are characterized by complex dynamics, making accurate modelling only viable through numerical simulation, and posing significant challenges for control design. This paper proposes a reinforcement learning (RL) control framework for HWWE systems, in which the real-time control policy is learned directly through interactions with high-fidelity simulation. A numerical model is established for a HWWE system consisting of an IEA 15 MW wind turbine, a VolturnUS semi-submersible platform, and three torus-type WECs, which is then employed as the RL training environment. Control performance is evaluated in terms of both wave energy generation and platform motion reduction, two competing objectives, from a Pareto perspective. It is shown that the proposed RL controller achieves substantial Pareto improvements over conventional control strategies, e.g., over 75\% higher wave energy capture at the same platform motion level, or nearly 50\% lower motion at the same energy capture level, thereby significantly extending the attainable performance boundary of HWWE systems. △ Less

Submitted 11 August, 2026; originally announced August 2026.

Transfer Learning-Enabled Distortion Compensation for Amplitude-Phase-Time Block Modulation-Based Nonlinear Single-Carrier Wireless Communications

Authors: Guoxing Duan, Min Fan, Cheng Yi, Bensheng Yang, Wei Xu, Haiming Wang, Xiaohu You

Abstract: Power amplifier (PA) nonlinearity and memory effects significantly limit the spectral compliance, reliability, and energy efficiency of communication systems. To address this, we propose a transfer-learning-enabled, fully digital transceiver-cooperative method for amplitude-phase-time block modulation (APTBM)-based nonlinear single-carrier transmission under adjacent channel leakage ratio (ACLR) c… ▽ More Power amplifier (PA) nonlinearity and memory effects significantly limit the spectral compliance, reliability, and energy efficiency of communication systems. To address this, we propose a transfer-learning-enabled, fully digital transceiver-cooperative method for amplitude-phase-time block modulation (APTBM)-based nonlinear single-carrier transmission under adjacent channel leakage ratio (ACLR) constraints. At the transmitter, iterative clipping and filtering (ICAF) and static digital pre-distortion (SDPD) act jointly to reduce signal peaks and suppress spectral regrowth without requiring wideband feedback. At the receiver, the inherent amplitude-phase constraints of APTBM provide weakly supervised prior knowledge for offline inverse-model pretraining, which is followed by the online few-shot adaptation of a lightweight digital post-distortion (DPoD) network. Subsequently, a cascaded DPoD and clipping-noise cancellation scheme systematically compensates for residual distortions induced by both the PA and ICAF. Simulation and measurement results demonstrate reliable transmission at an input back-off of approximately 2 dB under a 30-dBc ACLR constraint. Furthermore, the proposed DPoD approach significantly reduces online training time and computational overhead, delivering a performance gain of over 2 dB compared to conventional learning-based DPoD schemes. △ Less

Submitted 9 August, 2026; originally announced August 2026.

Comments: 13 pages, 14 figures, 1 table

Towards Autonomous Formulaic Alpha Discovery: An Evolutionary Computation Perspective

Authors: Xinwei Yu, Yiyang Fu, Mingcheng Fan, Enqi Li, Yilin Gao, Shugong Xu

Abstract: Automated formulaic alpha discovery aims to generate predictive and interpretable trading signals from large symbolic factor spaces. Its effectiveness is constrained by noisy fitness estimates, market nonstationarity, costly backtesting, semantic redundancy, and conflicting practical objectives. Existing studies employ diverse techniques, including genetic programming (GP), evolutionary algorithms… ▽ More Automated formulaic alpha discovery aims to generate predictive and interpretable trading signals from large symbolic factor spaces. Its effectiveness is constrained by noisy fitness estimates, market nonstationarity, costly backtesting, semantic redundancy, and conflicting practical objectives. Existing studies employ diverse techniques, including genetic programming (GP), evolutionary algorithms (EAs), reinforcement learning (RL), generative flow networks (GFlowNets), Monte Carlo tree search (MCTS), large language models (LLMs), and agentic workflows, but generally examine them as separate algorithmic families. This article introduces, for the first time, a unified evolutionary computation (EC) perspective on automated formulaic alpha discovery, formulating it as a noisy, dynamic, and multiobjective symbolic evolutionary optimization problem. A six-component analytical framework is developed to characterize existing methods through representation, variation, fitness evaluation, selection, memory, and adaptation. Furthermore, an eight-dimensional, autonomy-oriented evaluation framework is proposed, covering search efficiency, fitness reliability, residual alpha quality, economic diversity, tradability, evolutionary autonomy, robustness to nonstationarity, and reproducibility. Together, these frameworks provide a systematic foundation for unifying heterogeneous approaches, diagnosing component-level limitations, and guiding the development of reliable, adaptive, interpretable, and reproducible autonomous alpha discovery systems. △ Less

Submitted 3 August, 2026; originally announced August 2026.

Focus Is All You Need: Adaptive Goal-aware Attention Orchestration for Multi-Agent Graph Systems

Authors: Mingzhou Fan, Siyuan Xu, Mingxuan Yuan

Abstract: Large language models (LLMs) enable autonomous agents for reasoning, planning, and tool use. Recent systems increasingly organize these agents as graphs of specialized, interconnected nodes. Although graph-based orchestration supports flexible decomposition and coordination, it creates a key challenge: \textbf{attention allocation}. As workflows grow, existing approaches often execute graph compon… ▽ More Large language models (LLMs) enable autonomous agents for reasoning, planning, and tool use. Recent systems increasingly organize these agents as graphs of specialized, interconnected nodes. Although graph-based orchestration supports flexible decomposition and coordination, it creates a key challenge: \textbf{attention allocation}. As workflows grow, existing approaches often execute graph components uniformly, wasting resources on irrelevant or low-impact tasks. We introduce \textbf{Attention Orchestration}, a paradigm that extends Transformer-style attention from token representations to workflow-level agent coordination. Our framework, \textbf{Adaptive Goal-aware Attention Orchestration (AGAO)}, dynamically estimates agent importance based on user objectives, graph dependencies, and computational constraints. AGAO combines three components: (1) goal-aware attention, measuring semantic relevance between user goals and agent capabilities; (2) topology-aware attention, modeling structural dependencies in agent graphs; and (3) resource-aware attention, allocating budgets and execution priorities across heterogeneous agents. Together, these mechanisms transform static agent graphs into adaptive systems that focus computation on goal-critical reasoning paths. Experiments across diverse multi-agent workloads show that AGAO improves task effectiveness while reducing unnecessary computation, latency, and token consumption compared with existing graph-based execution strategies. Our work establishes \textbf{Attention Engineering} as a direction for scalable, intelligent multi-agent systems. Code: https://github.com/MingzhouFan97/AGAO. △ Less

Submitted 26 July, 2026; originally announced July 2026.

DAIS: Dependency-Aware Intermediate QA Supervision for Complex Reasoning

Authors: Yu Wang, Ming Fan, Xicheng Zhang, Zhiyong Li, Zhihu Wang, Caiyue Xu, Dahai Hu, Ting Liu

Abstract: Chain-of-thought (CoT) supervision exposes intermediate rationales, but flat rationale targets usually optimize a single reasoning sequence and provide limited supervision on how local conclusions should support later decisions. We introduce Dependency-Aware Intermediate QA Supervision (DAIS), a training-time framework that converts filtered teacher rationales into stage-level QA records. Each int… ▽ More Chain-of-thought (CoT) supervision exposes intermediate rationales, but flat rationale targets usually optimize a single reasoning sequence and provide limited supervision on how local conclusions should support later decisions. We introduce Dependency-Aware Intermediate QA Supervision (DAIS), a training-time framework that converts filtered teacher rationales into stage-level QA records. Each intermediate record predicts a local answer conditioned on the previous states needed for that decision, while the final-answer record keeps the original task format; evaluation therefore uses only the original input and optional context. Across GDPR, AIACT, MedQA, and FOLIO with multiple Qwen backbones, DAIS improves average final-answer accuracy over answer-only, flat chain-of-thought, and independent-QA baselines. On policy-compliance benchmarks, it achieves a largest gain of 5.6% and an average gain of 4.2% over the strongest non-DAIS baseline. Controlled ablations show that valid previous-state conditioning contributes beyond longer targets or additional intermediate text, supporting dependency-conditioned intermediate QA as a lightweight auxiliary supervision signal for standard final-answer inference. △ Less

Submitted 21 July, 2026; originally announced July 2026.

Tracing the Shadows: Automatic Tracking and Analysis of Crypto Money Laundering via Transaction Semantic Analysis

Authors: Hao Wu, Haijun Wang, Shangwang Li, Yin Wu, Ming Fan, Ting Liu, Xiapu Luo

Abstract: With the rapid advancement of decentralized finance (DeFi), security incidents related to cryptocurrency have become increasingly prevalent. After such incidents, attackers typically attempt to rapidly move stolen assets, concealing the origin of illicit funds and ultimately converting them into fiat currency. However, existing anti-money laundering (AML) methods struggle to cope with the semantic… ▽ More With the rapid advancement of decentralized finance (DeFi), security incidents related to cryptocurrency have become increasingly prevalent. After such incidents, attackers typically attempt to rapidly move stolen assets, concealing the origin of illicit funds and ultimately converting them into fiat currency. However, existing anti-money laundering (AML) methods struggle to cope with the semantic complexity of DeFi transactions. They either rely heavily on low-level token transfers, or perform protocol-agnostic money flow analysis, failing to capture the high-level intent of transactions. In this paper, we propose AMLGuard, a semantic-aware AML framework for account-based blockchains. AMLGuard tracks illicit fund flows from known malicious addresses by performing semantic analysis on complex DeFi transactions, enabling accurate and continuous laundering tracking. Given a complex transaction, AMLGuard combines static rule-based analysis with retrieval-augmented large language model (LLM) reasoning to infer implicit DeFi semantics, transforming raw transaction data into high-level semantic representations. Furthermore, for cross-chain transactions where laundering intent is not explicitly exposed, AMLGuard parses transaction parameters and performs argument parsing to recover cross-chain semantics, enabling seamless tracking across ledgers. Based on the inferred semantics, AMLGuard abstracts each transaction into a DeFi Semantic Unit (DSU). We evaluate the effectiveness of AMLGuard on 82 real-world laundering cases, involving illicit assets worth over $1 billion. Specifically, AMLGuard reconstructs compact illicit fund-flow topologies with destination precision of 94.4% and 87.6%, while achieving the highest address recall of 98.4% and 95.8% and destination recall of 94.1% and 93.8% on single-chain and cross-chain datasets. △ Less

Submitted 21 July, 2026; originally announced July 2026.

TrapHunter: Exposing Covert Pathways in Trap Token Contracts

Authors: Yin Wu, Yixuan Liu, Yi Li, Chenyang Peng, Hao Wu, Ming Fan, Ting Liu, Haijun Wang

Abstract: Standardized token contracts (e.g., ERC-20) form the foundation of digital assets. However, attackers increasingly abuse this standardization to disguise malicious trap tokens. Unlike obvious violations, these contracts employ a strategy of "deceptive adherence": they strictly adhere to standard protocols to evade detection while embedding covert logic to defraud users. To address this, we first s… ▽ More Standardized token contracts (e.g., ERC-20) form the foundation of digital assets. However, attackers increasingly abuse this standardization to disguise malicious trap tokens. Unlike obvious violations, these contracts employ a strategy of "deceptive adherence": they strictly adhere to standard protocols to evade detection while embedding covert logic to defraud users. To address this, we first systematize the trap landscape by proposing a novel taxonomy derived from the intrinsic functional lifecycle of tokens (Generation, Circulation, Persistence, and Observation). We then propose TrapHunter, a framework designed to identify these traps and expose covert pathways within these deceptive contracts via intent deviation analysis. Specifically, TrapHunter introduces a unified semantic representation combining Abstract Behavior Trees (ABTs) and Augmented Path Graphs (APGs) to normalize intra-procedural syntax and reveal hidden execution paths driven by inter-procedural state dependencies. Crucially, it bridges the semantic gap by leveraging LLMs to reason about the behavioral intent of deviations from reference implementations, followed by fork-based dynamic validation to confirm exploitability. Experimental evaluation on 269 real-world contracts with three LLMs (DeepSeek, GPT, and Gemini) demonstrates that TrapHunter effectively detects all six categories of traps, achieving an average precision of 81.8% and recall of 85.4%, significantly outperforming state-of-the-art tools. △ Less

Submitted 21 July, 2026; originally announced July 2026.

Comments: Accepted to ISSTA 2026

ACM Class: D.2.4; D.2.5; K.6.5

Journal ref: Proceedings of the ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA '26), October 03--09, 2026, Oakland, California

Hidden Decoding at Scale: Latent Computation Scaling for Large Language Models

Authors: Aiwei Liu, Cheng Shi, Chuhan Wu, Ci Lei, Di Lu, Donald He, Fan Zhang, Fanhao Kong, Feifei Zhang, Guan Wang, Haicheng Wang, Haoyu Liu, Houjin Yu, Jiachen Ding, Jiayi Feng, Jie Zhou, Jijun Chi, Jindi Shi, Jing Lei, Junjie Zhang, Laiyi Li, Le Tian, Linhao Zhang, Miao Fan, Sijun Zhang , et al. (23 additional authors not shown)

Abstract: Scaling Large Language Models (LLMs) has been driven mainly by enlarging the Transformer backbone, but for an already-strong model this requires another round of costly pretraining. We study whether an existing backbone can keep improving by allocating more computation to each token while leaving the Transformer backbone fixed. Depth-recurrent (looped) Transformers pursue this goal but are hard to… ▽ More Scaling Large Language Models (LLMs) has been driven mainly by enlarging the Transformer backbone, but for an already-strong model this requires another round of costly pretraining. We study whether an existing backbone can keep improving by allocating more computation to each token while leaving the Transformer backbone fixed. Depth-recurrent (looped) Transformers pursue this goal but are hard to scale, because looped computation does not fit naturally with the pipeline parallelism used to train the largest models. We add computation along the sequence-length dimension, where the extra computation is simply a longer input and stays compatible with standard large-model training. We propose Hidden Decoding, a sequence-length scaling method applied during continued pretraining (CPT). It expands each token into n streams with independent embedding tables and keeps the intermediate streams' key-value cache as context, so each token performs more internal computation without adding or widening Transformer layers. To keep this affordable at scale, we introduce Stream-Factorized Attention, in which most layers attend only within each stream and only a few layers mix across streams, reducing the attention cost from quadratic to roughly linear in n. Experiments support two scaling results. At frontier scale, we train WeLM-HD4-80B and WeLM-HD4-617B at n=4 and improve their matched non-HD baselines, making Hidden Decoding the first demonstrated sequence-length scaling method at the 100B+ MoE scale. Across expansion factors, the gains grow as n increases, showing that sequence-length expansion is a practical fixed-backbone scaling path for frontier-scale LLMs. △ Less

Submitted 9 July, 2026; originally announced July 2026.

Comments: 30 pages, 9 figures

MSC Class: 68T50 ACM Class: I.2.7

Progress- and Reliability-Oriented Group Policy Optimization for Agentic Reinforcement Learning

Authors: Mingxuan Fan, Peiyang Liu

Abstract: Group-based reinforcement learning (RL) has become an effective paradigm for improving large language model agents on long-horizon interactive tasks. To obtain finer-grained policy updates than trajectory-level optimization, recent work has moved toward step-level group-based RL, where intermediate steps are grouped and compared within a rollout batch. However, step-level advantage estimation is s… ▽ More Group-based reinforcement learning (RL) has become an effective paradigm for improving large language model agents on long-horizon interactive tasks. To obtain finer-grained policy updates than trajectory-level optimization, recent work has moved toward step-level group-based RL, where intermediate steps are grouped and compared within a rollout batch. However, step-level advantage estimation is sensitive to how groups are formed: grouping by broad state keys improves coverage but may compare actions taken under different histories, while enforcing historical consistency yields fairer comparisons at the cost of fragmented groups and missing peer-comparison signal. In this paper, we propose ProGPO (Progress- and Reliability-Oriented Group Policy Optimization), a learned-critic-free method for context-consistent step-level learning. ProGPO keeps exact-prefix action comparison, and complements sparse peer comparisons with transition credit derived from rollout-based state potentials. To estimate these potentials reliably, ProGPO combines semantic expansion with inverse-variance fusion across history depths. We evaluate ProGPO on two challenging agentic tasks, ALFWorld and WebShop, with Qwen2.5-1.5B-Instruct. Results show that ProGPO improves over matched agentic RL baselines under comparable computational overhead, and additional Qwen2.5-3B-Instruct experiments further test the scalability of the proposed method. △ Less

Submitted 5 July, 2026; originally announced July 2026.

Floor Raiser or Ceiling Limiter? Differential Storytelling Outcomes with a Child-Centric GenAI System Across Individual Differences

Authors: Min Fan, Wanqing Ma, Xinyue Cui, Xiaolu Dai, Shengyu Huang

Abstract: Generative AI (GenAI) holds promise for democratizing creative literacy, yet whether it benefits all children equally remains unclear. Using a child-centric GenAI storytelling system for children aged 7-12, we conducted a mixed-methods within-subjects experiment (N = 40, Grades 2-6) comparing GenAI-assisted and traditional storyboard conditions. Three findings emerged. First, the GenAI-assisted co… ▽ More Generative AI (GenAI) holds promise for democratizing creative literacy, yet whether it benefits all children equally remains unclear. Using a child-centric GenAI storytelling system for children aged 7-12, we conducted a mixed-methods within-subjects experiment (N = 40, Grades 2-6) comparing GenAI-assisted and traditional storyboard conditions. Three findings emerged. First, the GenAI-assisted condition was associated with a floor-raising convergence pattern, with the quality gap narrowing by 83.5%, driven by lower-end support and upper-end constraint mechanisms. This convergence was dimension-selective, improving creativity and richness while leaving coherence and narrative structure tied to baseline performance. Second, younger children more often selected semantically distant keywords while older children preferred semantically closer ones, although engagement orientation varied across individuals regardless of age. Third, image regeneration was positively associated with structural quality dimensions, though this association was attenuated after baseline control. We propose mechanism-contingent scaffolding as a design principle for adaptive GenAI storytelling systems serving diverse children. △ Less

Submitted 25 June, 2026; originally announced June 2026.

Towards Robust Personalized Federated Learning: Vulnerability Assessment and Defense Co-Design

Authors: Mingyuan Fan, Cen Chen

Abstract: The proliferation of IoT devices has fueled distributed edge systems to collect vast amounts of sensitive data, creating fertile ground for on-device machine learning applications. While federated learning (FL) mitigates privacy concerns by exchanging model parameters instead of raw data, we identify a critical blind spot in current research. We examine the most commonly used personalized federate… ▽ More The proliferation of IoT devices has fueled distributed edge systems to collect vast amounts of sensitive data, creating fertile ground for on-device machine learning applications. While federated learning (FL) mitigates privacy concerns by exchanging model parameters instead of raw data, we identify a critical blind spot in current research. We examine the most commonly used personalized federated learning (PFL) methods, which allow clients to maintain private, personalized models to address data heterogeneity across clients. Through systematic analysis, we reveal that PFL methods exhibit heightened vulnerability to transfer-based adversarial attacks compared to centralized learning paradigms. Wherein, malicious clients can exploit local model knowledge to craft adversarial examples that can compromise peer clients' personalized models. We establish this vulnerability through both theoretical analysis and empirical evaluation across multiple benchmark datasets, demonstrating significant accuracy drops across various PFL methods. To address this challenge, we propose a defense framework combining stochastic input noise, input-scaled trace regularization, and parameter sensitivity maximization to improve FL's robustness. Our findings establish the first systematic study of adversarial threats in PFL systems, providing both diagnostic tools and practical countermeasures. △ Less

Submitted 21 June, 2026; originally announced June 2026.

Amplitude-Phase-Frequency Block Modulation for OFDM-ISAC with SI-Free PAPR Reduction and Pilotless Sensing

Authors: Bensheng Yang, Min Fan, Haitao Zhao, Haiming Wang

Abstract: Orthogonal Frequency Division Multiplexing (OFDM)-based integrated sensing and communication systems demand a unified waveform that simultaneously supports reliable data transmission, low peak-to-average power ratio (PAPR), and accurate channel sensing. Existing approaches multiplex communication and sensing across separate time or frequency resources, or rely on dedicated pilots for channel estim… ▽ More Orthogonal Frequency Division Multiplexing (OFDM)-based integrated sensing and communication systems demand a unified waveform that simultaneously supports reliable data transmission, low peak-to-average power ratio (PAPR), and accurate channel sensing. Existing approaches multiplex communication and sensing across separate time or frequency resources, or rely on dedicated pilots for channel estimation, limiting system flexibility and increasing overhead. This paper proposes an amplitude-phase-frequency block modulation (APFBM) scheme for OFDM that achieves waveform-level integration of communication and sensing without resource partitioning. Information symbols are represented on the Stokes sphere and mapped to energy-normalized Jones vectors through an unambiguous rule that establishes a deterministic phase reference per block. This mapping exposes a commonphase degree of freedom inherent in the signal structure. At the transmitter, a grouped phase optimization algorithm exploits this structural freedom to reduce the PAPR without side information (SI). At the receiver, the same deterministic phase structure enables a Viterbi-based maximum-likelihood (ML) sequence detection algorithm that jointly recovers the optimization phases and estimates the block-wise channel amplitude and phase. No dedicated sensing pilots are required, as the sensing observables are extracted directly from the communication waveform. Closed-form error-rate and sensing-accuracy expressions are derived. Numerical simulations and over-the-air measurements on a software-defined radio link confirm effective PAPR reduction, accurate channel sensing, reliable phase recovery, and stable channel state information reconstruction. The proposed scheme trades a moderate reduction in spectral efficiency for a unified waveform design that simultaneously delivers SI-free PAPR reduction and pilotless sensing. △ Less

Submitted 18 June, 2026; originally announced June 2026.

CODA-BENCH: Can Code Agents Handle Data-Intensive Tasks?

Authors: Yuxin Zhang, Ju Fan, Meihao Fan, Shaolei Zhang, Xiaoyong Du

Abstract: Advanced agents are increasingly demonstrating the potential to operate as autonomous engineers, creating a growing demand for evaluation benchmarks that capture the complexity of real-world development. Such environments typically involve both complex code and large-scale data (i.e., file system). However, existing benchmarks usually evaluate code-centric or data-centric capabilities in isolation… ▽ More Advanced agents are increasingly demonstrating the potential to operate as autonomous engineers, creating a growing demand for evaluation benchmarks that capture the complexity of real-world development. Such environments typically involve both complex code and large-scale data (i.e., file system). However, existing benchmarks usually evaluate code-centric or data-centric capabilities in isolation, leaving a clear gap with real development scenarios. In this paper, we bridge this gap by introducing CODA-BENCH, the first benchmark to jointly evaluate code and data intelligence in a data-intensive environment. We construct a data-intensive Linux sandbox based on the Kaggle ecosystem (containing hundreds of datasets), where agents must actively explore complex file hierarchies to identify relevant resources and generate code for data-driven analytical tasks. CODA-BENCH comprises 1,009 tasks spanning 31 communities, with each task environment containing an average of 980 files, simulating realistic data scale and noise. Evaluations of advanced agents reveal that even top-performing systems struggle to effectively integrate data discovery with code execution, achieving a success rate of only 61.1%. These results highlight a substantial gap in current agentic capabilities for data-intensive tasks and point to promising directions for future research. △ Less

Submitted 13 June, 2026; originally announced June 2026.

Comments: Accepted at ICML 2026. 37 pages, 11 figures. Project page: https://coda-bench.github.io/ Code: https://github.com/ruc-datalab/CoDA-Bench Data: https://huggingface.co/datasets/RUC-DataLab/CoDA-Bench

Ling and Ring 2.6 Technical Report: Efficient and Instant Agentic Intelligence at Trillion-Parameter Scale

Authors: Ang Li, Ben Liu, Bin Han, Bin Hu, Bin Jing, Binbin Hu, Bing Li, Cai Chen, Caizhi Tang, Changxin Tian, Chao Huang, Chao Zhang, Chen Liang, Chen Qian, Chengfu Tang, Chengyao Wen, Chilin Fu, Chunwei Wu, Cong Zhang, Cunyin Peng, Daixin Wang, Dalong Zhang, Deng Zhao, Dingnan Jin, Dingyuan Zhu , et al. (193 additional authors not shown)

Abstract: Efficient and scalable agentic intelligence requires models that can deliver both low-latency responses and strong reasoning capabilities while remaining practical to train, serve, and deploy. In this report, we present Ling-2.6 and Ring-2.6, a family of models designed to address this challenge at scale. Ling-2.6 is optimized for instant response generation and high capability per output token, w… ▽ More Efficient and scalable agentic intelligence requires models that can deliver both low-latency responses and strong reasoning capabilities while remaining practical to train, serve, and deploy. In this report, we present Ling-2.6 and Ring-2.6, a family of models designed to address this challenge at scale. Ling-2.6 is optimized for instant response generation and high capability per output token, whereas Ring-2.6 is tailored for deeper reasoning and more advanced agentic workflows. Instead of training from scratch, we upgrade the Ling-2.0 base model through architectural migration pre-training and large-scale post-training. This upgrade is guided by a unified co-design of model architecture, optimization objectives, serving systems, and agent training environments, enabling improvements in both model capability and deployment efficiency. At the architectural level, we introduce a hybrid linear attention design that integrates Lightning Attention with MLA, improving the efficiency of long-context training and decoding. To further enhance token efficiency, we optimize capability per output token through Evolutionary Chain-of-Thought, Linguistic Unit Policy Optimization, bidirectional preference alignment, and shortest-correct-response distillation. For agentic capabilities, we propose KPop, a reinforcement learning framework designed to support stable training of Ring-2.6-1T on large-scale environment-grounded data. KPop improves training efficiency through asynchronous scheduling across coding, search, tool use, and workflow execution, enabling scalable learning from complex agent-environment interactions. Together, Ling-2.6 and Ring-2.6 provide a practical pathway toward efficient, scalable, and open agentic systems. We open-source all checkpoints in the 2.6 family to support further research and development in practical agentic intelligence. △ Less

Submitted 12 June, 2026; originally announced June 2026.

TACO: A Benchmark for Open-Domain Text-to-SQL with Ambiguous and Cross-Database Queries

Authors: Chao Deng, Ju Fan, Yuyu Luo, Qinliang Xue, Meihao Fan, Yuxin Zhang, Min Zhang, Xiaofeng Jia, Jing Zhang, Xiaoyong Du

Abstract: Text-to-SQL aims to translate natural language questions into executable SQL queries over structured databases. Existing benchmarks mainly focus on closed-domain settings with predefined database schemas and well-specified questions, but they fall short in addressing the challenges of open-domain scenarios, such as ambiguous questions, unspecified databases, and cross-database querying. To bridge… ▽ More Text-to-SQL aims to translate natural language questions into executable SQL queries over structured databases. Existing benchmarks mainly focus on closed-domain settings with predefined database schemas and well-specified questions, but they fall short in addressing the challenges of open-domain scenarios, such as ambiguous questions, unspecified databases, and cross-database querying. To bridge this gap, we introduce TACO, a benchmark for open-domain Text-to-SQL with Ambiguous and Cross-database queries. TACO consists of 1,500 real-world Text-to-SQL examples based on a smart city data service and 13,000 high-quality synthetic examples generated based on large-scale open data portals, covering diverse domains such as transportation, healthcare, and finance. To construct the synthetic examples, we develop an effective data synthesis pipeline that preserves the complexity of real-world queries. To demonstrate the utility of TACO, we introduce a baseline TACO-SQL composed of question rewriting, table linking, and query planning, to illustrate the challenges posed by TACO and to better understand the limitations of existing Text-to-SQL approaches. Extensive experiments on TACO using a variety of recent Text-to-SQL approaches show that, while TACO-SQL achieves the best results, a significant gap still remains between the existing approaches and human-written SQL. These findings highlight the difficulty of open-domain Text-to-SQL and position TACO as a valuable benchmark to drive future research. △ Less

Submitted 12 June, 2026; originally announced June 2026.

Limit on the nuclear Schiff moment of europium-153

Authors: Bassam Nima, Mingyu Fan, Xubo Wang, Sen Wang, En Fu Zhou, Andrew M. Jayich, Jiang Ming Yao, Lan Cheng, Amar Vutha

Abstract: The Schiff moment of a nucleus is a symmetry-violating nuclear moment that indicates new physics beyond the Standard Model. We place the limit, $|\mathscr{S}({}^{153}$Eu)$| < 1.7 \times 10^{-8}$ $e\,$fm$^3$ (95\% confidence), on the Schiff moment of the $^{153}$Eu nucleus, using nuclear spin resonances in two ensembles of oppositely-polarized $^{153}$Eu$^{3+}$ ions in a Y${}_2$SiO${}_5$ crystal. T… ▽ More The Schiff moment of a nucleus is a symmetry-violating nuclear moment that indicates new physics beyond the Standard Model. We place the limit, $|\mathscr{S}({}^{153}$Eu)$| < 1.7 \times 10^{-8}$ $e\,$fm$^3$ (95\% confidence), on the Schiff moment of the $^{153}$Eu nucleus, using nuclear spin resonances in two ensembles of oppositely-polarized $^{153}$Eu$^{3+}$ ions in a Y${}_2$SiO${}_5$ crystal. This measurement using octupolar nuclei in a mm-scale crystal constrains new physics at the TeV energy-scale. △ Less

Submitted 1 August, 2026; v1 submitted 10 June, 2026; originally announced June 2026.

Dominant in-plane anomalous Hall effect in a monoclinic room-temperature ferromagnet

Authors: Guoxin Zheng, Arjyama Bordoloi, Mingjun Fan, Shunsuke Kitou, Hiraku Saito, Taro Nakajima, Sobhit Singh, Takashi Kurumaji, Linda Ye

Abstract: Ferromagnetic metals are characterized by enhanced dissipationless transverse transport responses via the anomalous Hall effect, offering a route towards magnetic sensing and spintronic readout functionalities. In most ferromagnets, the anomalous Hall current is constrained to lie in the plane perpendicular to the magnetization (or applied magnetic field). Recently, it has been recognized that sel… ▽ More Ferromagnetic metals are characterized by enhanced dissipationless transverse transport responses via the anomalous Hall effect, offering a route towards magnetic sensing and spintronic readout functionalities. In most ferromagnets, the anomalous Hall current is constrained to lie in the plane perpendicular to the magnetization (or applied magnetic field). Recently, it has been recognized that selected symmetries can also permit a Hall response in a traditionally forbidden configuration, where the Hall current lies in the same plane as the magnetization, realizing an in-plane anomalous Hall effect. Reported realizations of this effect, however, are typically much weaker than the conventional Hall response in the same material. Here, through engineering specific crystallographic mirror symmetry-breaking, we realize a strongly enhanced in-plane anomalous Hall response in monoclinic Cr3Te4 with room-temperature ferromagnetism. Remarkably, the in-plane anomalous Hall signal exceeds the out-of-plane response by a factor of five, with which we demonstrate a unique in-plane field and current sensing functionality. Combined with density functional theory calculations, our results establish low-crystalline-symmetry ferromagnets with near-Fermi-level Weyl points as a practical platform for symmetry-engineered Hall responses, and point to a route towards room-temperature, geometry-flexible sensing devices. △ Less

Submitted 10 June, 2026; v1 submitted 8 June, 2026; originally announced June 2026.

TAGA: Terrain-aware Active Gaze Learning for Generalizable Agile Humanoid Locomotion

Authors: Peizhuo Li, Hongyi Li, Mingfeng Fan, Fangzhou Xu, Shuhao Liao, Yuxuan Ma, Zicheng Zeng, Ze Wang, Yongbin Jin, Yuhong Cao, Hongtao Wang, Guillaume Sartoretti

Abstract: Agile humanoid locomotion across diverse challenging terrain demands both wide perceptual coverage and precise local geometry understanding. Motivated by the way humans selectively look at relevant terrain during locomotion, we introduce TAGA, a Terrain-aware Active Gaze learning framework for Attention-based humanoid control. By fusing vision, proprioception, and motion commands, our framework gu… ▽ More Agile humanoid locomotion across diverse challenging terrain demands both wide perceptual coverage and precise local geometry understanding. Motivated by the way humans selectively look at relevant terrain during locomotion, we introduce TAGA, a Terrain-aware Active Gaze learning framework for Attention-based humanoid control. By fusing vision, proprioception, and motion commands, our framework guides the model to learn anticipatory cues and actively attend to specific areas of the height scan, selectively using these informative regions for the downstream network. This adaptively increases the information density of observations under tight onboard computational constraints, thus enabling fine-grained perceptive locomotion over larger-scale terrains. We find that such gaze behaviors can naturally emerge through reinforcement learning alone, without requiring additional supervision or explicit guidance, significantly improve training efficiency. As a result, the trained policy demonstrates robust and generalizable locomotion in simulation and on hardware, including reliable terrain-aware foothold selection, elevated-platform traversal, competitive sparse-foothold traversal, and the largest reported real-world gap traversal distance of 1.2m among perceptive humanoid locomotion systems, while maintaining stability under severe perceptual disturbances and environmental interference. △ Less

Submitted 4 June, 2026; originally announced June 2026.

Evaluating Interactive Reasoning in Large Language Models: A Hierarchical Benchmark with Executable Games

Authors: Mingyuan Fan, Weiguang Han, Daixin Wang, Cen Chen, Zhiqiang Zhang, Jun Zhou

Abstract: We introduce a multi-turn interactive framework for reasoning evaluation that treats reasoning as active evidence acquisition and belief updating. Wherein, LLMs receive only the task rules, must issue targeted queries to a hidden environment, integrate partial observations over time, and decide when to submit a final answer. Beyond standard success rate and interaction efficiency, we evaluate cont… ▽ More We introduce a multi-turn interactive framework for reasoning evaluation that treats reasoning as active evidence acquisition and belief updating. Wherein, LLMs receive only the task rules, must issue targeted queries to a hidden environment, integrate partial observations over time, and decide when to submit a final answer. Beyond standard success rate and interaction efficiency, we evaluate contextual robustness under controlled contextual perturbations, and metacognitive adaptation through counterfactual revision and necessity judgment. We instantiate the framework as a benchmark of 474 executable games, each evaluated under five fixed configuration search spaces corresponding to five difficulty levels, and evaluate a broad set of frontier LLMs. Results show that the benchmark is highly discriminative, exposing large differences not only in success rate but also in interaction efficiency. Moreover, we empirically show that contextual perturbations cause moderate but consistent declines, whereas counterfactual revision and necessity judgment lead to much larger drops. △ Less

Submitted 26 May, 2026; originally announced June 2026.

Comments: preprint version, under review

BitsMoE: Efficient Spectral Energy-Guided Bit Allocation for MoE LLM Quantization

Authors: Jiayu Zhao, Zihan Teng, Minhao Fan, Tianrui Ma, Wentao Ren, Song Chen, Weichen Liu

Abstract: Mixture-of-Experts (MoE) large language models reduce per-token computation through sparse expert activation, but their deployment remains memory-intensive because all expert weights must be kept resident in memory. Existing MoE compression methods struggle in the ultra-low-bit regime: pruning irreversibly removes model capacity, while coarse-grained quantization fails to allocate bits according t… ▽ More Mixture-of-Experts (MoE) large language models reduce per-token computation through sparse expert activation, but their deployment remains memory-intensive because all expert weights must be kept resident in memory. Existing MoE compression methods struggle in the ultra-low-bit regime: pruning irreversibly removes model capacity, while coarse-grained quantization fails to allocate bits according to heterogeneous expert and weight-direction importance. We propose BitsMoE, a spectral-energy-guided bit-allocation framework for MoE LLM quantization. BitsMoE decomposes each MoE layer by SVD into a shared basis and expert-specific spectral factors, retaining the shared basis without quantization to preserve common cross-expert structure and using the expert-specific factors as fine-grained quantization units. To determine the bit-width of each unit, BitsMoE formulates spectrum-wise mixed-precision quantization as an activation-aware reconstruction surrogate and solves an integer linear program that minimizes estimated reconstruction loss under a fixed bit budget. Experiments across multiple MoE LLMs show that BitsMoE substantially reduces downstream task accuracy degradation in ultra-low-bit regimes. Under 2-bit quantization on Qwen3-30B-A3B-Base, BitsMoE accelerates quantization by 12.3$\times$, improves average accuracy by 27.83 percentage points, and increases decoding speed by 1.76$\times$ over GPTQ. Our model and code are publicly available at https://github.com/zjiayu064/BitsMoE. △ Less

Submitted 22 May, 2026; originally announced June 2026.

Comments: 29 pages, 6 figures, 9 tables. Code and models are available at https://github.com/zjiayu064/BitsMoE

Testing the physical reality of tidal bulges in the world's oceans

Authors: Yongfeng Yang, Jiajia Yuan, Mingyuan Fan

Abstract: Persistent alternation of high and low water in coastal and oceanic regions has attracted human attention for millennia. This movement of water is generally explained through the double water bulge model. Although this model has been widely adopted in the scientific literature on tides since the 18th century, the physical existence of water bulges on the Earth's surface has yet to be verified. Her… ▽ More Persistent alternation of high and low water in coastal and oceanic regions has attracted human attention for millennia. This movement of water is generally explained through the double water bulge model. Although this model has been widely adopted in the scientific literature on tides since the 18th century, the physical existence of water bulges on the Earth's surface has yet to be verified. Herein, we establish a lunar angle phase-dependent statistical analysis of tide patterns at 362,370 oceanic locations spotted by Jason-3 satellite of AVISO in 2021 to address this issue. We show that during lunar angle phases of 0 degree-45 degree and 135 degree-180 degree, which spatially correspond to the water bulging regions expected in the double water bulge model, the number of low tides consistently exceeds that of high tides. Conversely, during lunar angle phase of 45 degree-135 degree, which spatially correspond to the water-depressing region expected in the model, high tides predominantly outnumber low tides. These findings evidently contradict the physical existence of two water bulges in the world's oceans, suggesting that the scientific community should pay additional attention to alternative explanations for tides, such as gravitational forcing mechanism and oceanic basin oscillation-generated driving mechanism. △ Less

Submitted 28 May, 2026; originally announced May 2026.

Comments: 18 pages, 6 figures, 1 table. Sci. China Earth Sci. (2026)

MeniOmni: A Structured Multimodal Benchmark for Holistic Meniscus Injury Assessment

Authors: Shurui Xu, Siqi Yang, Weiping Ding, Hui Wang, Mengzhen Fan, Yuyu Sun, Shuyan Li

Abstract: Clinical diagnosis of meniscus injuries requires radiologists to integrate volumetric MRI evidence with patient context (e.g., sex, age, BMI) and to produce structured diagnostic reports. Existing knee MRI benchmarks are typically unimodal and rely on coarse labels, limiting their ability to evaluate holistic clinical reasoning. We introduce MeniOmni, a structured multimodal benchmark for meniscus… ▽ More Clinical diagnosis of meniscus injuries requires radiologists to integrate volumetric MRI evidence with patient context (e.g., sex, age, BMI) and to produce structured diagnostic reports. Existing knee MRI benchmarks are typically unimodal and rely on coarse labels, limiting their ability to evaluate holistic clinical reasoning. We introduce MeniOmni, a structured multimodal benchmark for meniscus injury assessment, consisting of 746 multi-center MRI studies with tri-planar volumetric inputs, Clinical Priors, and expert-annotated clinical text. MeniOmni supports two tasks: (1) fine-grained Stoller severity grading and (2) diagnostic report generation. We further propose risk-aware ordinal evaluation and a semantic consistency metric (Meni-Score) to better reflect clinical relevance. Baseline experiments show that incorporating Clinical Priors improves grading performance and reduces severe errors, highlighting the value of multimodal context for safer assessment. Code and data are available at https://github.com/ShuruiXu/MeniOmni. △ Less

Submitted 27 May, 2026; originally announced May 2026.

Comments: Accepted by IEEE International Conference on Multimedia and Expo (ICME) 2026 (Oral Presentation)

Detail Consistent Stage-Wise Distillation for Efficient 3D MRI Segmentation

Authors: Mengchen Fan, Baocheng Geng, Xi Xiao, Tianyang Wang, Siyuan Mei, Pulin Che, Xiaoqian Jiang, Qizhen Lan

Abstract: Deploying high-performing 3D medical image segmenters (e.g., nnU-Net) is often limited by memory footprint and inference latency. Compression is therefore necessary, but compact 3D encoders tend to lose fine structural cues (small lesions and sharp boundaries) as downsampling repeats across multi-resolution stages. We propose Detail Consistent Distillation (DCD), a stage-wise distillation framewor… ▽ More Deploying high-performing 3D medical image segmenters (e.g., nnU-Net) is often limited by memory footprint and inference latency. Compression is therefore necessary, but compact 3D encoders tend to lose fine structural cues (small lesions and sharp boundaries) as downsampling repeats across multi-resolution stages. We propose Detail Consistent Distillation (DCD), a stage-wise distillation framework that preserves structural detail across scales by aligning teacher-student features in a wavelet-decomposed representation. At each encoder stage, DCD distills directional detail components in the wavelet domain while leaving the coarse approximation comparatively unconstrained, avoiding over-regularization of global semantics. DCD is used only during training and introduces no inference-time overhead. Experiments on the BraTS 2024 and ISLES 2022 benchmarks demonstrate that our approach achieves superior performance in MRI segmentation using 3D multi-modal data. Code and implementation details for DCD are publicly available at https://github.com/ClinicaAlpha/DCD-3D-MedSeg. △ Less

Submitted 25 May, 2026; originally announced May 2026.

Comments: Accepted by MICCAI 2026. 11 pages, 3 figures

MemAudit: Post-hoc Auditing of Poisoned Agent Memory via Causal Attribution and Structural Anomaly Detection

Authors: Zhewen Tan, Yilun Yao, Huiyan Jin, Wenhan Yu, Guoan Wang, Mengyuan Fan, liang lu, Feng Liu, Xiangzheng Zhang, Duohe Ma, Tong Yang, Lin Sun

Abstract: Large language model agents increasingly rely on persistent memory to store past interactions, retrieve relevant demonstrations, and improve long-horizon task execution. However, this memory mechanism also creates a practical security vulnerability: an adversarial user may inject malicious records into the agent's memory through ordinary interaction, and these records can later be retrieved to ste… ▽ More Large language model agents increasingly rely on persistent memory to store past interactions, retrieve relevant demonstrations, and improve long-horizon task execution. However, this memory mechanism also creates a practical security vulnerability: an adversarial user may inject malicious records into the agent's memory through ordinary interaction, and these records can later be retrieved to steer the agent's reasoning and actions. Existing defenses primarily focus on online intervention, such as prompt filtering or output blocking, but they do not address the post-hoc question of which stored memories are responsible after harmful behavior has already been observed. We propose \textbf{MemAudit}, a post-hoc causal memory auditing framework for memory-augmented LLM agents. The framework combines two complementary signals: (1) a counterfactual memory influence score that measures each memory's causal contribution to harmful outputs, and (2) a memory consistency graph that identifies structurally anomalous memories within the broader memory store. We evaluate MemAudit against MINJA, a query-only memory injection attack in which malicious records are generated and stored through normal agent interactions rather than direct memory-bank modification. Across both QA and reasoning-agent settings, MemAudit substantially reduces attack success rates under realistic post-hoc auditing scenarios. The results show that QA attack success is reduced from $70\%$ to $0\%$, while RAP attack success drops from $83.3\%$ to $0\%$. △ Less

Submitted 22 May, 2026; originally announced May 2026.

What Does the Server See? Understanding Privacy Leakage from Large Language Models in Split Inference

Authors: Mingyuan Fan, Yu Liu, Fuyi Wang, Cen Chen

Abstract: The deployment of large language models (LLMs) on resource-constrained devices remains challenging, spurring interest in split inference, where models are partitioned between client and server to reduce computational burden and enhance privacy by transmitting only intermediate activations. However, the privacy-preserving capabilities of split inference, particularly in the context of LLMs, have no… ▽ More The deployment of large language models (LLMs) on resource-constrained devices remains challenging, spurring interest in split inference, where models are partitioned between client and server to reduce computational burden and enhance privacy by transmitting only intermediate activations. However, the privacy-preserving capabilities of split inference, particularly in the context of LLMs, have not been exhaustively investigated. To fill this gap, we introduce ActInv, which solves an intermediate activation matching problem to reconstruct the client's input. Extensive evaluations demonstrate that ActInv achieves high-fidelity reconstructions, even in the presence of common perturbation-based defenses such as Gaussian noise injection and activation sparsification. To systematically understand this vulnerability, we develop Perturbation Amplification Factor (PAF), a metric for quantifying a layer's inherent resistance to reconstruction. Our analysis reveals that privacy vulnerability is not uniform across layers, with some layers being highly susceptible to leakage while others offer natural resistance. Furthermore, we demonstrate that defense effectiveness can be significantly improved by calibrating perturbation directions to maximize reconstruction error during backpropagation. Building on these insights, we design PriPert and conduct comprehensive evaluations, covering privacy, utility, and computational overhead, to demonstrate its effectiveness. △ Less

Submitted 21 May, 2026; originally announced May 2026.

Comments: Accepted to ACM CCS'26

Learning to Solve Compositional Geometry Routing Problems

Authors: Mingfeng Fan, Jianan Zhou, Jiaqi Cheng, Yifeng Zhang, Jie Zhang, Guillaume Adrien Sartoretti

Abstract: We study the Compositional Geometry Routing Problem (CGRP), a unified superclass of traditional routing problems that covers point-only, line-only, area-only, and arbitrary hybrid task geometries, providing a broad abstraction for real-world routing scenarios. Beyond standard point-based routing, CGRP with non-point tasks can be inherently asymmetric, tightly coupled travel routes with the intrins… ▽ More We study the Compositional Geometry Routing Problem (CGRP), a unified superclass of traditional routing problems that covers point-only, line-only, area-only, and arbitrary hybrid task geometries, providing a broad abstraction for real-world routing scenarios. Beyond standard point-based routing, CGRP with non-point tasks can be inherently asymmetric, tightly coupled travel routes with the intrinsic path, and enlarges the action space with numerous feasible yet often irrelevant options, thereby posing significant challenges for both representation learning and decision-making. To address these challenges, we propose DiCon, a differential attention-assisted solver with contrastive learning, as a plug-and-play framework that tackles the problem from two complementary angles. First, we introduce a differential attention mechanism that actively suppresses the probability mass on less competitive candidate actions. Second, we design a double-level contrastive learning objective to promote robust global instance representations and regularize geometry-aware task representations. Extensive experiments demonstrate that DiCon achieves strong performance, broad versatility, and superior generalization across diverse CGRP instances with different compositions. △ Less

Submitted 18 May, 2026; originally announced May 2026.

Comments: 27 pages, 10 figures

Beyond Compliance: How AI Could Help Creative Writers by Refusing Them

Authors: Hua Xuan Qin, Guangzhi Zhu, Mingming Fan, Pan Hui

Abstract: Mainstream creativity support design prioritizes compliant AI for seamless writing interactions, but concerns over inappropriate AI reliance highlight the need for designs fostering reflection on balanced AI and non-AI resource use. Theoretically, intentional AI non-compliance, refusals (saying ``no'' to requests), could introduce such reflection through friction stronger than other bypass-able so… ▽ More Mainstream creativity support design prioritizes compliant AI for seamless writing interactions, but concerns over inappropriate AI reliance highlight the need for designs fostering reflection on balanced AI and non-AI resource use. Theoretically, intentional AI non-compliance, refusals (saying ``no'' to requests), could introduce such reflection through friction stronger than other bypass-able solutions. Practically, refusal content/language characteristics lead to nuanced reactions. However, little research empirically focuses on nuances beyond mandatory ethical/technical constraints, on turning refusals into strategic friction for `innocuous' requests. We address this through a qualitative study with 22 creative writers, exploring reactions to refusals to common requests across writing stages (planning, translating, reviewing). Findings suggest that reflective potential depends on heterogeneous preference alignment along situational (e.g., convergent/divergent thinking phases), cognitive (e.g., domain beliefs), and relational (e.g., AI roles) dimensions. We discuss implications for creativity support, broader issues (e.g., AI addiction), and frictional/seamful AI design (e.g., integrating different compliance levels). △ Less

Submitted 3 April, 2026; originally announced May 2026.

Comments: conditionally accepted to Creativity & Cognition 2026

MultiEmo-Bench: Multi-label Visual Emotion Analysis for Multi-modal Large Language Models

Authors: Tianwei Chen, Takuya Furusawa, Yuki Hirakawa, Ryotaro Shimizu, Mo Fan, Takashi Wada

Abstract: This paper introduces a multi-label visual emotion analysis benchmark dataset for comprehensively evaluating the ability of multimodal large language models (MLLMs) to predict the emotions evoked by images. Recent user studies report an unintuitive finding: humans may prefer the predictions of MLLMs over the labels in existing datasets. We argue that this phenomenon stems from the suboptimal annot… ▽ More This paper introduces a multi-label visual emotion analysis benchmark dataset for comprehensively evaluating the ability of multimodal large language models (MLLMs) to predict the emotions evoked by images. Recent user studies report an unintuitive finding: humans may prefer the predictions of MLLMs over the labels in existing datasets. We argue that this phenomenon stems from the suboptimal annotation scheme used in existing datasets, where each annotator is shown a single candidate emotion for each image and judges whether it is evoked or not. This approach is clearly limited because a single image can evoke multiple emotions with varying intensities. As a result, evaluations based on these datasets may underestimate the capabilities of MLLMs, yet an appropriate benchmark for evaluating such models remains lacking. To address this issue, we introduce a new multi-label benchmark dataset for visual emotion analysis toward MLLMs evaluation. We hire $20$ annotators per image and ask them to select all emotions they feel from an image. Then, we aggregate the votes across all annotators, providing a more reliable and representative dataset labeled with a distribution of emotions. The resulting dataset contains $10,344$ images with $236,998$ valid votes across eight emotions. Based on this benchmark dataset, we evaluate several recent models, including Qwen3-VL, OpenAI's GPT, Gemini, and Claude. We assess model performance on both dominant emotion prediction and emotion distribution prediction. Our results demonstrate the progress achieved by recent MLLMs while also indicating that substantial room for improvement remains. Furthermore, our experiments with LLM-as-a-judge show that the method does not consistently improve MLLMs' performance, indicating its limitations for the subjective task of visual emotion analysis. △ Less

Submitted 14 May, 2026; originally announced May 2026.

Stealthy Backdoor Attacks against LLMs Based on Natural Style Triggers

Authors: Jiali Wei, Ming Fan, Guoheng Sun, Xicheng Zhang, Haijun Wang, Ting Liu

Abstract: The growing application of large language models (LLMs) in safety-critical domains has raised urgent concerns about their security. Many recent studies have demonstrated the feasibility of backdoor attacks against LLMs. However, existing methods suffer from three key shortcomings: explicit trigger patterns that compromise naturalness, unreliable injection of attacker-specified payloads in long-for… ▽ More The growing application of large language models (LLMs) in safety-critical domains has raised urgent concerns about their security. Many recent studies have demonstrated the feasibility of backdoor attacks against LLMs. However, existing methods suffer from three key shortcomings: explicit trigger patterns that compromise naturalness, unreliable injection of attacker-specified payloads in long-form generation, and incompletely specified threat models that obscure how backdoors are delivered and activated in practice. To address these gaps, we present BadStyle, a complete backdoor attack framework and pipeline. BadStyle leverages an LLM as a poisoned sample generator to construct natural and stealthy poisoned samples that carry imperceptible style-level triggers while preserving semantics and fluency. To stabilize payload injection during fine-tuning, we design an auxiliary target loss that reinforces the attacker-specified target content in responses to poisoned inputs and penalizes its emergence in benign responses. We further ground the attack in a realistic threat model and systematically evaluate BadStyle under both prompt-induced and PEFT-based injection strategies. Extensive experiments across seven victim LLMs, including LLaMA, Phi, DeepSeek, and GPT series, demonstrate that BadStyle achieves high attack success rates (ASRs) while maintaining strong stealthiness. The proposed auxiliary target loss substantially improves the stability of backdoor activation, yielding an average ASR improvement of around 30% across style-level triggers. Even in downstream deployment scenarios unknown during injection, the implanted backdoor remains effective. Moreover, BadStyle consistently evades representative input-level defenses and bypasses output-level defenses through simple camouflage. △ Less

Submitted 23 April, 2026; originally announced April 2026.

Multi-View Hierarchical Graph Neural Network for Sketch-Based 3D Shape Retrieval

Authors: Hang Cheng, Muyan He, Mingyu Fan, Chengfeng Xie, Xi Cheng, Long Zeng

Abstract: Sketch-based 3D shape retrieval (SBSR) aims to retrieve 3D shapes that are consistent with the category of the input hand-drawn sketch. The core challenge of this task lies in two aspects: existing methods typically employ simplified aggregation strategies for independently encoded 3D multi-view features, which ignore the geometric relationships between views and multi-level details, resulting in… ▽ More Sketch-based 3D shape retrieval (SBSR) aims to retrieve 3D shapes that are consistent with the category of the input hand-drawn sketch. The core challenge of this task lies in two aspects: existing methods typically employ simplified aggregation strategies for independently encoded 3D multi-view features, which ignore the geometric relationships between views and multi-level details, resulting in weak 3D representation. Simultaneously, traditional SBSR methods are constrained by visible category limitations, leading to poor performance in zero-shot scenarios. To address these challenges, we propose Multi-View Hierarchical Graph Neural Network (MV-HGNN), a novel framework for SBSR. Specifically, we construct a view-level graph and capture adjacent geometric dependencies and cross-view message passing via local graph convolution and global attention. A view selector is further introduced to perform hierarchical graph coarsening, enabling a progressively larger receptive field for graph convolution and mitigating the interference of redundant views, which leads to more discriminate discriminative hierarchical 3D representation. To enable category agnostic alignment and mitigate overfitting to seen classes, we leverage CLIP text embeddings as semantic prototypes and project both sketch and 3D features into a shared semantic space. We use a two-stage training strategy for category-level retrieval and a one-stage strategy for zero-shot retrieval under the same model architecture. Under both category-level and zero-shot settings, extensive experiments on two public benchmarks demonstrate that MV-HGNN outperforms state-of-the-art methods. △ Less

Submitted 20 April, 2026; originally announced April 2026.

Momentum-constrained Hybrid Heuristic Trajectory Optimization Framework with Residual-enhanced DRL for Visually Impaired Scenarios

Authors: Yuting Zeng, Zhiwen Zheng, Jingya Wang, You Zhou, JiaLing Xiao, Yongbin Yu, Manping Fan, Bo Gong, Liyong Ren

Abstract: Safe and efficient assistive planning for visually impaired scenarios remains challenging, since existing methods struggle with multi-objective optimization, generalization, and interpretability. In response, this paper proposes a Momentum-Constrained Hybrid Heuristic Trajectory Optimization Framework (MHHTOF). To balance multiple objectives of comfort and safety, the framework designs a Heuristic… ▽ More Safe and efficient assistive planning for visually impaired scenarios remains challenging, since existing methods struggle with multi-objective optimization, generalization, and interpretability. In response, this paper proposes a Momentum-Constrained Hybrid Heuristic Trajectory Optimization Framework (MHHTOF). To balance multiple objectives of comfort and safety, the framework designs a Heuristic Trajectory Sampling Cluster (HTSC) with a Momentum-Constrained Trajectory Optimization (MTO), which suppresses abrupt velocity and acceleration changes. In addition, a novel residual-enhanced deep reinforcement learning (DRL) module refines candidate trajectories, advancing temporal modeling and policy generalization. Finally, a dual-stage cost modeling mechanism (DCMM) is introduced to regulate optimization, where costs in the Frenet space ensure consistency, and reward-driven adaptive weights in the Cartesian space integrate user preferences for interpretability and user-centric decision-making. Experimental results show that the proposed framework converges in nearly half the iterations of baselines and achieves lower and more stable costs. In complex dynamic scenarios, MHHTOF further demonstrates stable velocity and acceleration curves with reduced risk, confirming its advantages in robustness, safety, and efficiency. △ Less

Submitted 16 April, 2026; originally announced April 2026.

Comments: 24 pages, 14 figures. arXiv admin note: text overlap with arXiv:2509.15582

Enhance Comprehension of Over-the-Counter Drug Instructions for the General Public and Medical Professionals through Visualization Design

Authors: Mengjie Fan, Katrin Angerbauer, Yinchu Cheng, Yingying Yan, Xiaohan Xu, Tianfu Wang, Michael Sedlmair, Yu Yang, Liang Zhou

Abstract: Drug instructions are crucial for guiding the rational use of medication. We conduct a visualization design study to enhance the comprehension of over-the-counter (OTC) drug instructions, targeting both the general public and medical professionals. We devise two tailored drug instruction designs for different audience groups through an iterative design process. A controlled user study reveals that… ▽ More Drug instructions are crucial for guiding the rational use of medication. We conduct a visualization design study to enhance the comprehension of over-the-counter (OTC) drug instructions, targeting both the general public and medical professionals. We devise two tailored drug instruction designs for different audience groups through an iterative design process. A controlled user study reveals that our design outperforms traditional text-based instructions in terms of response time and usability, and the availability of two versions is also found to be beneficial. This study also motivates a taxonomy based on a systematic classification of OTC drug instructions sampled from an official drug database, which received positive expert feedback. Finally, this study summarizes a workflow for a visualization design strategy based on our design exploration and user study feedback, which can be generalized to other OTC drug instructions. △ Less

Submitted 10 April, 2026; originally announced April 2026.

Journal ref: Computers & Graphics, Volume 136, May 2026, 104587

TypePro: Boosting LLM-Based Type Inference via Inter-Procedural Slicing

Authors: Teyu Lin, Minghao Fan, Huaxun Huang, Zhirong Shen, Rongxin Wu

Abstract: Dynamic languages (such as Python and JavaScript) offer flexibility and simplified type handling for programming, but this can also lead to an increase in type-related errors and additional overhead for compile-time type inference. As a result, type inference for dynamic languages has become a popular research area. Existing approaches typically achieve type inference through static analysis, mach… ▽ More Dynamic languages (such as Python and JavaScript) offer flexibility and simplified type handling for programming, but this can also lead to an increase in type-related errors and additional overhead for compile-time type inference. As a result, type inference for dynamic languages has become a popular research area. Existing approaches typically achieve type inference through static analysis, machine learning, or large language models (LLMs). However, current work only focuses on the direct dependencies of variables related to type inference as the context, resulting in incomplete contextual information and thus affecting the accuracy of type inference. To address this issue, this paper proposes a method called TypePro, which leverages LLMs for type inference in dynamic languages. TypePro supplements contextual information by conducting inter-procedural code slicing. Then, TypePro proposes a set of candidate complex types based on the structural information of data types implied in the slices, thereby addressing the lack of domain knowledge of LLMs. We conducted experiments on the ManyTypes4Py and ManyTypes4TypeScript datasets, achieving Top-1 exact match (EM) rates of 88.9% and 86.6%, respectively. Notably, TypePro improves the Top-1 Exact Match by 7.1 and 10.3 percentage points over the second-best approach, showing the effectiveness and robustness of TypePro. △ Less

Submitted 2 April, 2026; originally announced April 2026.

CROSS: A Mixture-of-Experts Reinforcement Learning Framework for Generalizable Large-Scale Traffic Signal Control

Authors: Xibei Chen, Yifeng Zhang, Yuxiang Xiao, Mingfeng Fan, Maonan Wang, Guillaume Sartoretti

Abstract: Recent advances in robotics, automation, and artificial intelligence have enabled urban traffic systems to operate with increasing autonomy towards future smart cities, powered in part by the development of adaptive traffic signal control (ATSC), which dynamically optimizes signal phases to mitigate congestion and optimize traffic. However, achieving effective and generalizable large-scale ATSC re… ▽ More Recent advances in robotics, automation, and artificial intelligence have enabled urban traffic systems to operate with increasing autonomy towards future smart cities, powered in part by the development of adaptive traffic signal control (ATSC), which dynamically optimizes signal phases to mitigate congestion and optimize traffic. However, achieving effective and generalizable large-scale ATSC remains a significant challenge due to the diverse intersection topologies and highly dynamic, complex traffic demand patterns across the network. Existing RL-based methods typically use a single shared policy for all scenarios, whose limited representational capacity makes it difficult to capture diverse traffic dynamics and generalize to unseen environments. To address these challenges, we propose CROSS, a novel Mixture-of-Experts (MoE)-based decentralized RL framework for generalizable ATSC. We first introduce a Predictive Contrastive Clustering (PCC) module that forecasts short-term state transitions to identify latent traffic patterns, followed by clustering and contrastive learning to enhance pattern-level representation. We further design a Scenario-Adaptive MoE module that augments a shared policy with multiple experts, thus enabling adaptive specialization and more flexible scenario-specific strategies. We conduct extensive experiments in the SUMO simulator on both synthetic and real-world traffic datasets. Compared with state-of-the-art baselines, CROSS achieves superior performance and generalization through improved representation of diverse traffic scenarios. △ Less

Submitted 25 March, 2026; originally announced March 2026.

Sensing T-violating nuclear moments of paramagnetic ions in crystals

Authors: Aleksandar Radak, Mingyu Fan, Bassam Nima, Yuiki Takahashi, Amar Vutha

Abstract: Precision measurements of time-reversal (T) symmetry violating moments probe physics beyond the Standard Model. We show that precision spectroscopy of paramagnetic lanthanide and actinide ions doped into noncentrosymmetric crystals offers a promising platform for extending the sensitivity of searches for T-violation in nuclear physics. The unpaired valence electron in these ions allows the enginee… ▽ More Precision measurements of time-reversal (T) symmetry violating moments probe physics beyond the Standard Model. We show that precision spectroscopy of paramagnetic lanthanide and actinide ions doped into noncentrosymmetric crystals offers a promising platform for extending the sensitivity of searches for T-violation in nuclear physics. The unpaired valence electron in these ions allows the engineering of highly-coherent hyperfine transitions that are insensitive to magnetic fields, yet highly sensitive to new physics. These systems also provide other advantages for new physics searches, including large numbers of ions that can be measured in doped crystals, strong electric polarization of the ions by the crystal fields, enhancement of T-violating nuclear moments in nonspherical nuclei, and accurate comagnetometers generated by crystal symmetry. We estimate the new physics sensitivity of these proposed systems to be two orders of magnitude better than existing constraints. △ Less

Submitted 25 March, 2026; originally announced March 2026.

LATS: Large Language Model Assisted Teacher-Student Framework for Multi-Agent Reinforcement Learning in Traffic Signal Control

Authors: Yifeng Zhang, Peizhuo Li, Tingguang Zhou, Mingfeng Fan, Guillaume Sartoretti

Abstract: Adaptive Traffic Signal Control (ATSC) aims to optimize traffic flow and minimize delays by adjusting traffic lights in real time. Recent advances in Multi-agent Reinforcement Learning (MARL) have shown promise for ATSC, yet existing approaches still suffer from limited representational capacity, often leading to suboptimal performance and poor generalization in complex and dynamic traffic environ… ▽ More Adaptive Traffic Signal Control (ATSC) aims to optimize traffic flow and minimize delays by adjusting traffic lights in real time. Recent advances in Multi-agent Reinforcement Learning (MARL) have shown promise for ATSC, yet existing approaches still suffer from limited representational capacity, often leading to suboptimal performance and poor generalization in complex and dynamic traffic environments. On the other hand, Large Language Models (LLMs) excel at semantic representation, reasoning, and analysis, yet their propensity for hallucination and slow inference speeds often hinder their direct application to decision-making tasks. To address these challenges, we propose a novel learning paradigm named LATS that integrates LLMs and MARL, leveraging the former's strong prior knowledge and inductive abilities to enhance the latter's decision-making process. Specifically, we introduce a plug-and-play teacher-student learning module, where a trained embedding LLM serves as a teacher to generate rich semantic features that capture each intersection's topology structures and traffic dynamics. A much simpler (student) neural network then learns to emulate these features through knowledge distillation in the latent space, enabling the final model to operate independently from the LLM for downstream use in the RL decision-making process. This integration significantly enhances the overall model's representational capacity across diverse traffic scenarios, thus leading to more efficient and generalizable control strategies. Extensive experiments across diverse traffic datasets empirically demonstrate that our method enhances the representation learning capability of RL models, thereby leading to improved overall performance and generalization over both traditional RL and LLM-only approaches. [...] △ Less

Submitted 25 March, 2026; originally announced March 2026.

Blow-up of solutions to the Euler-Poisson-Darbox equation with critical power nonlinearity

Authors: Mengting Fan, Ning-An Lai, Hiroyuki Takamura

Abstract: In our recent precious work, we established the finite time blow up result and upper bound of lifespan estimate to the singular Cauchy problem of semilinear Euler-Poisson-Darboux equation in R^n with subcritical power type nonlinearity. By introducing an improved test function, we obtain an enhanced lower bound for the functional including the spacetime integral of the nonlinear term with an addit… ▽ More In our recent precious work, we established the finite time blow up result and upper bound of lifespan estimate to the singular Cauchy problem of semilinear Euler-Poisson-Darboux equation in R^n with subcritical power type nonlinearity. By introducing an improved test function, we obtain an enhanced lower bound for the functional including the spacetime integral of the nonlinear term with an additional logarithmic growth, which finally yields the blow up result and upper bound of lifespan estimate for the corresponding Cauchy problem with "critical" nonlinear power. And this gives some partial answer to the open problem 1 posed by D'Abbicco (J. Differential Equations 286 (2021), 531-556). △ Less

Submitted 26 March, 2026; v1 submitted 19 March, 2026; originally announced March 2026.

Comments: 26 pages. The dropped key references, [16, 17], related to (66) on p.15 are added in version 2

MSC Class: primary: 35L71; secondary: 35Q05

CAMO: A Conditional Neural Solver for the Multi-objective Multiple Traveling Salesman Problem

Authors: Fengxiaoxiao Li, Xiao Mao, Mingfeng Fan, Yifeng Zhang, Yi Li, Tanishq Duhan, Guillaume Sartoretti

Abstract: Robotic systems often require a team of robots to collectively visit multiple targets while optimizing competing objectives, such as total travel cost and makespan. This setting can be formulated as the Multi-Objective Multiple Traveling Salesman Problem (MOMTSP). Although learning-based methods have shown strong performance on the single-agent TSP and multi-objective TSP variants, they rarely add… ▽ More Robotic systems often require a team of robots to collectively visit multiple targets while optimizing competing objectives, such as total travel cost and makespan. This setting can be formulated as the Multi-Objective Multiple Traveling Salesman Problem (MOMTSP). Although learning-based methods have shown strong performance on the single-agent TSP and multi-objective TSP variants, they rarely address the combined challenges of multi-agent coordination and multi-objective trade-offs, which introduce dual sources of complexity. To bridge this gap, we propose CAMO, a conditional neural solver for MOMTSP that generalizes across varying numbers of targets, agents, and preference vectors, and yields high-quality approximations to the Pareto front (PF). Specifically, CAMO consists of a conditional encoder to fuse preferences into instance representations, enabling explicit control over multi-objective trade-offs, and a collaborative decoder that coordinates all agents by alternating agent selection and node selection to construct multi-agent tours autoregressively. To further improve generalization, we train CAMO with a REINFORCE-based objective over a mixed distribution of problem sizes. Extensive experiments show that CAMO outperforms both neural and conventional heuristics, achieving a closer approximation of PFs. In addition, ablation results validate the contributions of CAMO's key components, and real-world tests on a mobile robot platform demonstrate its practical applicability. △ Less

Submitted 19 March, 2026; originally announced March 2026.

Comments: 9 pages, 3 figures

"It Became My Buddy, But I'm Not Afraid to Disagree": A Multi-Session Study of UX Evaluators Collaborating with Conversational AI Assistants

Authors: Emily Kuang, Ehsan Jahangirzadeh Soure, Luyao Shen, Nitesh Goyal, Mingming Fan, Kristen Shinohara

Abstract: AI-assisted usability analysis can potentially reduce the time and effort of finding usability problems, yet little is known about how AI's perceived expertise influences evaluators' analytic strategies and perceptions over time. We ran a within-subjects, five-session study (six hours per participant) with 12 professional UX evaluators who worked with two conversational assistants designed to appe… ▽ More AI-assisted usability analysis can potentially reduce the time and effort of finding usability problems, yet little is known about how AI's perceived expertise influences evaluators' analytic strategies and perceptions over time. We ran a within-subjects, five-session study (six hours per participant) with 12 professional UX evaluators who worked with two conversational assistants designed to appear novice- or expert-like (differing in suggestion quantity and response accuracy). We logged behavioral measures (number of passes, suggestion acceptance rate), collected subjective ratings (trust, perceived efficiency), and conducted semi-structured interviews. Participants experienced an initial novelty effect and a subsequent dip in trust that recovered over time. Their efficiency improved as they shifted from a two-pass to a one-pass video inspection approach. Evaluators ultimately rated the experienced CA as significantly more efficient, trustworthy, and comprehensive, despite not perceiving expertise differences early on. We conclude with design implications for adapting AI expertise to enable calibrated human-AI collaboration. △ Less

Submitted 13 March, 2026; originally announced March 2026.

Comments: Accepted to CHI 2026

Recover to Predict: Progressive Retrospective Learning for Variable-Length Trajectory Prediction

Authors: Hao Zhou, Lu Qi, Jason Li, Jie Zhang, Yi Liu, Xu Yang, Mingyu Fan, Fei Luo

Abstract: Trajectory prediction is critical for autonomous driving, enabling safe and efficient planning in dense, dynamic traffic. Most existing methods optimize prediction accuracy under fixed-length observations. However, real-world driving often yields variable-length, incomplete observations, posing a challenge to these methods. A common strategy is to directly map features from incomplete observations… ▽ More Trajectory prediction is critical for autonomous driving, enabling safe and efficient planning in dense, dynamic traffic. Most existing methods optimize prediction accuracy under fixed-length observations. However, real-world driving often yields variable-length, incomplete observations, posing a challenge to these methods. A common strategy is to directly map features from incomplete observations to those from complete ones. This one-shot mapping, however, struggles to learn accurate representations for short trajectories due to significant information gaps. To address this issue, we propose a Progressive Retrospective Framework (PRF), which gradually aligns features from incomplete observations with those from complete ones via a cascade of retrospective units. Each unit consists of a Retrospective Distillation Module (RDM) and a Retrospective Prediction Module (RPM), where RDM distills features and RPM recovers previous timesteps using the distilled features. Moreover, we propose a Rolling-Start Training Strategy (RSTS) that enhances data efficiency during PRF training. PRF is plug-and-play with existing methods. Extensive experiments on datasets Argoverse 2 and Argoverse 1 demonstrate the effectiveness of PRF. Code is available at https://github.com/zhouhao94/PRF. △ Less

Submitted 11 March, 2026; originally announced March 2026.

Comments: Paper is accepted by CVPR 2026

TaPD: Temporal-adaptive Progressive Distillation for Observation-Adaptive Trajectory Forecasting in Autonomous Driving

Authors: Mingyu Fan, Yi Liu, Hao Zhou, Deheng Qian, Mohammad Haziq Khan, Matthias Raetsch

Abstract: Trajectory prediction is essential for autonomous driving, enabling vehicles to anticipate the motion of surrounding agents to support safe planning. However, most existing predictors assume fixed-length histories and suffer substantial performance degradation when observations are variable or extremely short in real-world settings (e.g., due to occlusion or a limited sensing range). We propose Ta… ▽ More Trajectory prediction is essential for autonomous driving, enabling vehicles to anticipate the motion of surrounding agents to support safe planning. However, most existing predictors assume fixed-length histories and suffer substantial performance degradation when observations are variable or extremely short in real-world settings (e.g., due to occlusion or a limited sensing range). We propose TaPD (Temporal-adaptive Progressive Distillation), a unified plug-and-play framework for observation-adaptive trajectory forecasting under variable history lengths. TaPD comprises two cooperative modules: an Observation-Adaptive Forecaster (OAF) for future prediction and a Temporal Backfilling Module (TBM) for explicit reconstruction of the past. OAF is built on progressive knowledge distillation (PKD), which transfers motion pattern knowledge from long-horizon "teachers" to short-horizon "students" via hierarchical feature regression, enabling short observations to recover richer motion context. We further introduce a cosine-annealed distillation weighting scheme to balance forecasting supervision and feature alignment, improving optimization stability and cross-length consistency. For extremely short histories where implicit alignment is insufficient, TBM backfills missing historical segments conditioned on scene evolution, producing context-rich trajectories that strengthen PKD and thereby improve OAF. We employ a decoupled pretrain-reconstruct-finetune protocol to preserve real-motion priors while adapting to backfilled inputs. Extensive experiments on Argoverse 1 and Argoverse 2 show that TaPD consistently outperforms strong baselines across all observation lengths, delivers especially large gains under very short inputs, and improves other predictors (e.g., HiVT) in a plug-and-play manner. Code will be available at https://github.com/zhouhao94/TaPD. △ Less

Submitted 6 March, 2026; originally announced March 2026.

Fed-GAME: Personalized Federated Learning with Graph Attention Mixture-of-Experts For Time-Series Forecasting

Authors: Yi Li, Han Liu, Mingfeng Fan, Guo Chen, Chaojie Li, Biplab Sikdar

Abstract: Federated learning (FL) on graphs shows promise for distributed time-series forecasting. Yet, existing methods rely on static topologies and struggle with client heterogeneity. We propose Fed-GAME, a framework that models personalized aggregation as message passing over a learnable dynamic implicit graph. The core is a decoupled parameter difference-based update protocol, where clients transmit pa… ▽ More Federated learning (FL) on graphs shows promise for distributed time-series forecasting. Yet, existing methods rely on static topologies and struggle with client heterogeneity. We propose Fed-GAME, a framework that models personalized aggregation as message passing over a learnable dynamic implicit graph. The core is a decoupled parameter difference-based update protocol, where clients transmit parameter differences between their fine-tuned private model and a shared global model. On the server, these differences are decomposed into two streams: (1) averaged difference used to updating the global model for consensus (2) the selective difference fed into a novel Graph Attention Mixture-of-Experts (GAME) aggregator for fine-grained personalization. In this aggregator, shared experts provide scoring signals while personalized gates adaptively weight selective updates to support personalized aggregation. Experiments on two real-world electric vehicle charging datasets demonstrate that Fed-GAME outperforms state-of-the-art personalized FL baselines. △ Less

Submitted 1 March, 2026; originally announced March 2026.

SkyReels-V4: Multi-modal Video-Audio Generation, Inpainting and Editing model

Authors: Guibin Chen, Dixuan Lin, Jiangping Yang, Youqiang Zhang, Zhengcong Fei, Debang Li, Sheng Chen, Chaofeng Ao, Nuo Pang, Yiming Wang, Yikun Dou, Zheng Chen, Mingyuan Fan, Tuanhui Li, Mingshan Chang, Hao Zhang, Xiaopeng Sun, Jingtao Xu, Yuqiang Xie, Jiahua Wang, Zhiheng Xu, Weiming Xiong, Yuzhe Jin, Baoxuan Gu, Binjie Mao , et al. (26 additional authors not shown)

Abstract: SkyReels V4 is a unified multi modal video foundation model for joint video audio generation, inpainting, and editing. The model adopts a dual stream Multimodal Diffusion Transformer (MMDiT) architecture, where one branch synthesizes video and the other generates temporally aligned audio, while sharing a powerful text encoder based on the Multimodal Large Language Models (MLLM). SkyReels V4 accept… ▽ More SkyReels V4 is a unified multi modal video foundation model for joint video audio generation, inpainting, and editing. The model adopts a dual stream Multimodal Diffusion Transformer (MMDiT) architecture, where one branch synthesizes video and the other generates temporally aligned audio, while sharing a powerful text encoder based on the Multimodal Large Language Models (MLLM). SkyReels V4 accepts rich multi modal instructions, including text, images, video clips, masks, and audio references. By combining the MLLMs multi modal instruction following capability with in context learning in the video branch MMDiT, the model can inject fine grained visual guidance under complex conditioning, while the audio branch MMDiT simultaneously leverages audio references to guide sound generation. On the video side, we adopt a channel concatenation formulation that unifies a wide range of inpainting style tasks, such as image to video, video extension, and video editing under a single interface, and naturally extends to vision referenced inpainting and editing via multi modal prompts. SkyReels V4 supports up to 1080p resolution, 32 FPS, and 15 second duration, enabling high fidelity, multi shot, cinema level video generation with synchronized audio. To make such high resolution, long-duration generation computationally feasible, we introduce an efficiency strategy: Joint generation of low resolution full sequences and high-resolution keyframes, followed by dedicated super-resolution and frame interpolation models. To our knowledge, SkyReels V4 is the first video foundation model that simultaneously supports multi-modal input, joint video audio generation, and a unified treatment of generation, inpainting, and editing, while maintaining strong efficiency and quality at cinematic resolutions and durations. △ Less

Submitted 18 March, 2026; v1 submitted 25 February, 2026; originally announced February 2026.

Trajectory Generation with Endpoint Regulation and Momentum-Aware Dynamics for Visually Impaired Scenarios

Authors: Yuting Zeng, Manping Fan, You Zhou, Yongbin Yu, Zhiwen Zheng, Jingtao Zhang, Liyong Ren, Zhenglin Yang

Abstract: Trajectory generation for visually impaired scenarios requires smooth and temporally consistent state in structured, low-speed dynamic environments. However, traditional jerk-based heuristic trajectory sampling with independent segment generation and conventional smoothness penalties often lead to unstable terminal behavior and state discontinuities under frequent regenerating. This paper proposes… ▽ More Trajectory generation for visually impaired scenarios requires smooth and temporally consistent state in structured, low-speed dynamic environments. However, traditional jerk-based heuristic trajectory sampling with independent segment generation and conventional smoothness penalties often lead to unstable terminal behavior and state discontinuities under frequent regenerating. This paper proposes a trajectory generation approach that integrates endpoint regulation to stabilize terminal states within each segment and momentum-aware dynamics to regularize the evolution of velocity and acceleration for segment consistency. Endpoint regulation is incorporated into trajectory sampling to stabilize terminal behavior, while a momentum-aware dynamics enforces consistent velocity and acceleration evolution across consecutive trajectory segments. Experimental results demonstrate reduced acceleration peaks and lower jerk levels with decreased dispersion, smoother velocity and acceleration profiles, more stable endpoint distributions, and fewer infeasible trajectory candidates compared with a baseline planner. △ Less

Submitted 25 February, 2026; originally announced February 2026.

Comments: 9 pages, 7 figures

Adaptive Multi-Scale Channel-Spatial Attention Aggregation Framework for 3D Indoor Semantic Scene Completion Toward Assisting Visually Impaired

Authors: Qi He, XiangXiang Wang, Jingtao Zhang, Yongbin Yu, Hongxiang Chu, Manping Fan, JingYe Cai, Zhenglin Yang

Abstract: Independent indoor mobility remains a critical challenge for individuals with visual impairments, largely due to the limited capability of existing assistive systems in detecting fine-grained hazardous objects such as chairs, tables, and small obstacles. These perceptual blind zones substantially increase the risk of collision in unfamiliar environments. To bridge the gap between monocular 3D visi… ▽ More Independent indoor mobility remains a critical challenge for individuals with visual impairments, largely due to the limited capability of existing assistive systems in detecting fine-grained hazardous objects such as chairs, tables, and small obstacles. These perceptual blind zones substantially increase the risk of collision in unfamiliar environments. To bridge the gap between monocular 3D vision research and practical assistive deployment, this paper proposes an Adaptive Multi-scale Attention Aggregation (AMAA) framework for monocular 3D semantic scene completion using only a wearable RGB camera. The proposed framework addresses two major limitations in 2D-to-3D feature lifting: noise diffusion during back-projection and structural instability in multi-scale fusion. A parallel channel--spatial attention mechanism is introduced to recalibrate lifted features along semantic and geometric dimensions, while a hierarchical adaptive gating strategy regulates cross-scale information flow to preserve fine-grained structural details. Experiments on the NYUv2 benchmark demonstrate that AMAA achieves an overall mIoU of 27.88%. Crucially, it yields significant relative improvements of 16.9% for small objects and 10.4% for tables over the MonoScene baseline. Furthermore, a wearable prototype based on an NVIDIA Jetson Orin NX and a ZED~2i camera validates stable real-time performance in indoor environments, demonstrating the feasibility of deploying monocular 3D scene completion for assistive navigation. △ Less

Submitted 15 April, 2026; v1 submitted 18 February, 2026; originally announced February 2026.

Comments: We need to optimize the experiment, the changes are quite significant

Time-independent theoretical framework for stroboscopic nonlinear dynamics based on time-nonlocal response

Authors: Yuhui Zhuang, Jiaxin Li, Haidong Li, Siyu Li, Xiaobin Peng, Juan Wu, Jiameng Zhang, Mingjing Fan, Xiaoqin Huang, Yi Hu, Jingjun Xu

Abstract: Recent experiments have demonstrated the ability to manipulate nonlinear interactions via time modulation, giving rise to the so-called stroboscopic nonlinearity. To date, however, this phenomenon has not been subjected to a rigorous theoretical analysis. In this work, we clarify the physical mechanism underlying stroboscopic nonlinear dynamics based on time-nonlocal response and establish an effe… ▽ More Recent experiments have demonstrated the ability to manipulate nonlinear interactions via time modulation, giving rise to the so-called stroboscopic nonlinearity. To date, however, this phenomenon has not been subjected to a rigorous theoretical analysis. In this work, we clarify the physical mechanism underlying stroboscopic nonlinear dynamics based on time-nonlocal response and establish an effective time-independent model under suitable modulation conditions. The proposed model almost exactly reproduces the full time-dependent dynamics in the quasi-steady state and significantly outperforms empirical descriptions used previously. Our results provide a clear physical picture of stroboscopic nonlinear dynamics, and can be extended to other systems with time-nonlocal response, establishing a general framework for engineering nonlinear interactions through temporal modulation. △ Less

Submitted 20 June, 2026; v1 submitted 7 February, 2026; originally announced February 2026.

DeepPrep: An LLM-Powered Agentic System for Autonomous Data Preparation

Authors: Meihao Fan, Ju Fan, Yuxin Zhang, Shaolei Zhang, Xiaoyong Du, Jie Song, Peng Li, Fuxin Jiang, Tieying Zhang, Jianjun Chen

Abstract: Data preparation, which aims to transform heterogeneous and noisy raw tables into analysis-ready data, remains a major bottleneck in data science. Recent approaches leverage large language models (LLMs) to automate data preparation from natural language specifications. However, existing LLM-powered methods either make decisions without grounding in intermediate execution results, or rely on linear… ▽ More Data preparation, which aims to transform heterogeneous and noisy raw tables into analysis-ready data, remains a major bottleneck in data science. Recent approaches leverage large language models (LLMs) to automate data preparation from natural language specifications. However, existing LLM-powered methods either make decisions without grounding in intermediate execution results, or rely on linear interaction processes that offer limited support for revising earlier decisions. To address these limitations, we propose DeepPrep, an LLM-powered agentic system for autonomous data preparation. DeepPrep constructs data preparation pipelines through iterative, execution-grounded interaction with an environment that materializes intermediate table states and returns runtime feedback. To overcome the limitations of linear interaction, DeepPrep organizes pipeline construction with tree-based agentic reasoning, enabling structured exploration and non-local revision based on execution feedback. To enable effective learning of such behaviors, we propose a progressive agentic training framework, together with data synthesis that supplies diverse and complex ADP tasks. Extensive experiments show that DeepPrep achieves data preparation accuracy comparable to strong closed-source models (e.g., GPT-5) while incurring 15x lower inference cost, while establishing state-of-the-art performance among open-source baselines and generalizing effectively across diverse datasets. △ Less

Submitted 7 February, 2026; originally announced February 2026.

ROMAN: Reward-Orchestrated Multi-Head Attention Network for Autonomous Driving System Testing

Authors: Jianlei Chi, Yuzhen Wu, Jiaxuan Hou, Xiaodong Zhang, Ming Fan, Suhui Sun, Weijun Dai, Bo Li, Jianguo Sun, Jun Sun

Abstract: Automated Driving System (ADS) acts as the brain of autonomous vehicles, responsible for their safety and efficiency. Safe deployment requires thorough testing in diverse real-world scenarios and compliance with traffic laws like speed limits, signal obedience, and right-of-way rules. Violations like running red lights or speeding pose severe safety risks. However, current testing approaches face… ▽ More Automated Driving System (ADS) acts as the brain of autonomous vehicles, responsible for their safety and efficiency. Safe deployment requires thorough testing in diverse real-world scenarios and compliance with traffic laws like speed limits, signal obedience, and right-of-way rules. Violations like running red lights or speeding pose severe safety risks. However, current testing approaches face significant challenges: limited ability to generate complex and high-risk law-breaking scenarios, and failing to account for complex interactions involving multiple vehicles and critical situations. To address these challenges, we propose ROMAN, a novel scenario generation approach for ADS testing that combines a multi-head attention network with a traffic law weighting mechanism. ROMAN is designed to generate high-risk violation scenarios to enable more thorough and targeted ADS evaluation. The multi-head attention mechanism models interactions among vehicles, traffic signals, and other factors. The traffic law weighting mechanism implements a workflow that leverages an LLM-based risk weighting module to evaluate violations based on the two dimensions of severity and occurrence. We have evaluated ROMAN by testing the Baidu Apollo ADS within the CARLA simulation platform and conducting extensive experiments to measure its performance. Experimental results demonstrate that ROMAN surpassed state-of-the-art tools ABLE and LawBreaker by achieving 7.91% higher average violation count than ABLE and 55.96% higher than LawBreaker, while also maintaining greater scenario diversity. In addition, only ROMAN successfully generated violation scenarios for every clause of the input traffic laws, enabling it to identify more high-risk violations than existing approaches. △ Less

Submitted 5 February, 2026; originally announced February 2026.

Comments: The manuscript includes 13 pages, 8 tables, and 7 figures

Direct vs. Score-based Selection: Understanding the Heisenberg Effect in Target Acquisition Across Input Modalities in Virtual Reality

Authors: Linjie Qiu, Duotun Wang, Boyu Li, Jiawei Li, Yulin Shen, Zeyu Wang, Mingming Fan

Abstract: Target selection is a fundamental interaction in virtual reality (VR). But the act of confirming a selection, such as a button press or pinch, can disturb the tracked pose and shift the intended target, which is referred to as the Heisenberg Effect. Prior research has mainly investigated controller input. However, it remains unclear how the effect manifests in the bare-hand input and how score-bas… ▽ More Target selection is a fundamental interaction in virtual reality (VR). But the act of confirming a selection, such as a button press or pinch, can disturb the tracked pose and shift the intended target, which is referred to as the Heisenberg Effect. Prior research has mainly investigated controller input. However, it remains unclear how the effect manifests in the bare-hand input and how score-based techniques may mitigate the effect in different spatial variations. To fill the gap, we conduct a within-subject study to examine the Heisenberg Effect across two input modalities (i.e., controller and hand) and two selection mechanisms (i.e., direct and score-based). Our results show that hand input is more susceptible to the Heisenberg Effect, with direct selection more influenced by target width and score-based selection more sensitive to target density. Based on previous vote-oriented technique and our temporal analysis, we introduce weighted VOTE, a history-based intention accuracy model for target voting, that reweights recent interaction intent to counteract input disturbances. Our evaluation shows the method improves selection accuracy compared to baseline techniques. Finally, we discuss future directions for adaptive selection methods. △ Less

Submitted 1 February, 2026; originally announced February 2026.

Comments: Accepted by TVCG and IEEE VR'26