arXiv subjects

Tao Chen

Publications and source records attributed to Tao Chen.

At least 55 records · Page 3Linked to original sources

arXiv2026

On AI Safety and Security Technical Debt in Engineering AI-Enabled Systems

Artificial intelligence (AI) systems are increasingly deployed in high-stakes domains such as healthcare, autonomous driving, finance, and education. While these systems offer powerful data-driven and adaptive capabilities, their complexity, rapid evolution, and dependence on dynamic data pipelines introduce new forms of engineering liability collectively referred to as AI Technical Debts (AITDs). AITDs arise from root causes spanning data governance, model implementation, algorithm design, architectural decisions, operational processes, documentation practices, and testing adequacy. Unlike conventional technical debt, many AITDs are latent and propagate across tightly coupled AI pipelines, leading to maintenance challenges, reliability degradation, and heightened safety or security risks. Guided by the principles of AI Trust, Risk, and Security Management (AI TRiSM), this study reinterprets technical debt through the interconnected dimensions of trustworthiness, focusing on AI safety and security technical debts. We conduct a systematic review of 60 primary studies and identify 31 distinct types of AITD, which are organized into a root-cause-oriented taxonomy comprising seven classes. The analysis examines how these debts map to 18 trust-related concerns, including 6 safety hazards and 12 security vulnerabilities. To support mitigation, the review synthesizes 34 actionable guidelines (8 safety and 26 security) targeting the prevention, detection, and reduction of AITDs across the AI lifecycle. Building on these findings, we introduce AITD-MAP, an integrated framework that connects the AITD taxonomy, quality and risk impacts, and mitigation strategies into a unified structure for risk-aware AI engineering. The framework aims to assist AI software engineers in making AI safety and security technical debts visible, understanding their root causes, and mitigating their presence.

cs.SE↗

arXiv2026

Token-Operations-Oriented Inference Optimization Techniques for Large Models

Large model inference optimization serves as a key foundation for supporting the scalable, low-cost, and highly stable operation of large model services. Centered on token-oriented inference optimization technology, this paper proposes for the first time a four-layer technical architecture consisting of Multi-model Fusion, Model Optimization, Compute-Model Fusion, and Compute-Network-Model Fusion. It systematically reviews the key technologies and current industry status across these four levels and analyzes the application value of related technologies in real-world business scenarios. This paper provides a practical technical path for reducing token production costs, improving token service efficiency, ensuring the stability of token supply, and driving the transition of large model services from being merely callable to being operable.

cs.SE↗

arXiv2026

MoE$^2$-LoRA: When MoE Models Meet MoE-style Low-Rank Adaptation

Mixture-of-Experts (MoE) architectures have been widely adopted in large language models, yet parameter-efficient fine-tuning (PEFT) for MoE models remains underexplored. Existing PEFT methods for MoE either ignore router priors with uniform adapters, reducing efficiency and risking forgetting, or rely on static expert selection, limiting per-token capacity and cross-expert feature learning. In this paper, we make the first attempt to fine-tune MoE models with MoE-style low-rank adaptation: our method, entitled MoE$^2$-LoRA, deeply couples the pretrained expert specialization with task-specific adaptivity via a dual-channel Routing-Conditioned Projection (RCP) module, which reuses base router activations to inform LoRA routing. We further introduce a single global LoRA expert pool shared across all layers, enabling model-wide adaptation with emergent layer-wise affinities and balanced expert utilization. MoE$^2$-LoRA simultaneously benefits from the advantages of prior reuse, dynamic adapter routing, and model-wide knowledge sharing. Evaluated on multiple MoE backbones with varying scales and expert granularities, MoE$^2$-LoRA consistently achieves state-of-the-art downstream accuracy while retaining stronger general capabilities.

cs.CL↗

arXiv2026

Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills

LLM training is shifting from manual design and annotation to interaction-driven self-evolution. However, existing self-evolutionary methods face a fundamental dilemma between task diversity and verification reliability: environment-bound methods obtain precise feedback but confine learning to narrow domains, while open-ended self-generation broadens the task space but lacks reliable verification, allowing misleading rewards to pollute the training loop. We identify agent skills as a powerful middle ground to reconcile this tension: each skill ensures deep, verifiable execution in a specific scenario, while dynamic routing across skills maintains open-ended task variety. Leveraging this insight, we introduce Skill Self-Play (Skill-SP), a co-evolutionary framework comprising a proposer, a solver, and a dynamic skill controller. Orchestrated via a reinforcement learning loop, these components co-evolve in a continuous self-play loop: the proposer generates challenging tasks conditioned on dynamically sampled skills; the solver explores candidate solutions to push its capability boundaries; and the skill controller collects execution feedback to update and expand the skill library. This interactive co-evolution effectively bridges the gap between structured verification and open-ended exploration. Empirical evaluations on tool-use and reasoning benchmarks demonstrate that Skill-SP, serving as a robust evolution engine, consistently pushes the performance ceiling of competent backbones while catalyzing striking turnarounds for initially misaligned models. Our code is available at https://github.com/Qwen-Applications/skill-self-play.

cs.CL↗

arXiv2026

Anharmonic Infrared Emission of Cyano-Substituted Polycyclic Aromatic Hydrocarbon Molecules: Cyanonaphthalenes as a Case Study

Recent detections of cyano-substituted polycyclic aromatic hydrocarbons (cyano-PAHs) highlight them as key tracers of nitrogen heterocycles in the interstellar medium (ISM). However, their infrared (IR) identification requires anharmonic vibrational radiative models that account for dynamic cooling across diverse environments. We provide IR cascade emission spectra for neutral, cationic, and anionic 1- and 2-cyanonaphthalene (1- and 2-CNN) under representative astrophysical conditions, investigating how charge states and substitution positions modulate anharmonic profiles and energy redistribution during IR cascades. Using B3LYP/N07D with Second-order Vibrational Perturbation Theory (VPT2), we computed anharmonic properties and applied an optimized microcanonical sampling algorithm to calculate vibrational density of states and model environment-dependent cascade spectra. Results show that charge states strongly tune the CN nitrile band: anion intensity is enhanced by an order of magnitude compared to neutrals, whereas cation intensity is slightly weaker. 1- and 2-CNN present distinct isomeric signatures: 1-CNN displays a complex "red-wing" structure in the 3.3-micron C-H stretch due to peri-hydrogen interactions, whereas 2-CNN yields a more symmetric profile. We evaluated the fractional energy emitted via the CN stretch relative to total cascade emissivity across charge states and environments. Combined with observed CN fluxes, these fractions provide a quantitative tool to constrain cyano-PAH abundances. By bridging laboratory benchmarks with dynamic interstellar emission, this work delivers accurate anharmonic fingerprints to guide JWST observational analysis.

astro-ph.GA↗

arXiv2026

Orbit-resolved spin holography: role of Coulomb focusing in target-dependent polarization

Strong-field photoelectron holography encodes ultrafast electron dynamics through momentum-space interference. However, the orbit-resolved origin of spider-like spin fringes and the mechanism underlying their target dependence remain unclear. Here, we resolve both issues by analyzing photoelectron spin textures generated during tunneling ionization. We use the Coulomb quantum-orbit strong-field approximation, benchmarked against time-dependent Schrödinger equation simulations for $\mathrm{He^+}$ and Xe, to separate orbital-channel and quantum-orbit contributions. Spider-like fringes arise from interference between $p$-orbital ionization channels with different magnetic quantum numbers within an individual orbit class and therefore do not require interorbit interference. The observable polarization along these fringes, however, depends on the balance among orbit-class contributions. The decomposition associates the opposite first-leg polarizations of $\mathrm{He^+}$ and Xe with different relative weights of laser-deflected and forward-scattered trajectories, consistent with target-dependent Coulomb focusing. Photoelectron spin textures thus complement momentum distributions as probes of Coulomb-driven strong-field dynamics.

physics.atom-ph↗

arXiv2026

Comparative CFD modelling of the interplay between aqueous humour hydrodynamics and drug and nanocarrier transport for glaucoma drug delivery via intracameral injection, drug-eluting implants and contact lenses

Glaucoma is the leading cause of irreversible blindness worldwide and often requires long-term administration of intraocular pressure-lowering agents. The effectiveness of ocular drug delivery depends not only on the delivery route but also on the transport mechanisms governing drug distribution within the eye. In this study, a computational fluid dynamics (CFD) framework was developed to investigate the influence of aqueous humour hydrodynamics on the transport of dissolved drugs and nanocarriers delivered by three clinically relevant strategies: intracameral injection, drug-eluting implants and contact lenses. An idealised three-dimensional model of the anterior segment was implemented in COMSOL Multiphysics v6.4, incorporating aqueous humour flow, thermal convection, drug transport, saccadic eye movements and particle tracing. This framework enabled direct comparison of molecular diffusion- and particle-mediated delivery. Simulations revealed substantial differences in intraocular transport among the delivery strategies. Contact-lens delivery produced the most homogeneous drug distributions, whereas implant-based delivery generated persistent concentration gradients associated with localised release. Intracameral injection exhibited intermediate behaviour, with rapid redistribution by aqueous humour circulation. Nanocarrier transport showed greater spatial heterogeneity than dissolved-drug transport owing to reduced diffusivity and consequently stronger advection-dominated transport. These findings demonstrate that aqueous humour hydrodynamics are an independent determinant of ocular drug transport and should be considered in the design of sustained ophthalmic drug delivery systems. The proposed framework provides a versatile computational tool for the optimisation of both dissolved-drug and nanocarrier-based glaucoma therapies.

q-bio.TO↗

arXiv2026

When to Use Which? Benchmarking Optimisers for Configurable Systems under Varying Budgets

Software configuration tuning is crucial for optimising system performance, and various optimisers have emerged over the last decade. Yet, the time required during the tuning process may vary across systems. In some systems (e.g., PostgreSQL), it may take a few minutes to measure a configuration, whereas in some others (e.g., MariaDB), it can take several hours. Moreover, even within the same system, users may have varying budgets and preferred settings. This naturally raises a question -- Given a budget level, which optimiser is the best choice for SE practitioners? This matters because optimisers usually have their own ``comfort zone'' and may perform very differently under distinct budgets. In this paper, we aim to answer this question. We systematically evaluate eight well-established optimisers across 22 configurable systems under varying budget levels. We find that, unsurprisingly, model-based optimisers (e.g., SMAC) are well-suited under tight budgets, and model-free optimisers (e.g., GAs) become superior with more generous budgets. However, interestingly, there is one optimiser, FLASH, that performs consistently well on most systems regardless of budgets. We lastly investigate the reasons behind this phenomenon and find that many systems possess good local optima (with large basins of attraction), allowing greedy optimisers (e.g., FLASH) to achieve strong performance. Source code, data, and supplementary materials of this work are available at https://anonymous.4open.science/r/Config-W2W-98B2.

cs.SE↗

arXiv2026

Optimal slit width for high-precision orbital angular momentum measurement using angular double-slit interferometry

We demonstrate an optimization of angular double-slit interferometry for accurate measurement of orbital angular momentum (OAM) of vortex beams. By scanning the dynamic double slits, the topological charge (TC) magnitude is directly determined from the oscillation frequency of the on-axis intensity. Based on repeated experimental investigations, we establish a critical criterion for slit width selection: to avoid phase truncation or period overlap, the angular width of each slit must exactly match the spiral phase period 2π/|l|. Under this optimal condition, the interference pattern exhibits the highest visibility and the measurement error is minimized. Experiments for l = 5, 10, and 15 are performed as representative examples, and the universality of this criterion is confirmed. Furthermore, by introducing an additional phase shift, the sign of the TC is unambiguously determined, as demonstrated for l = 10 and l = -15. This simple, robust method provides a high-precision pathway for OAM metrology.

physics.optics↗

arXiv2026

The existence and stability for a class of fuzzy fractional boundary value problems constrained by variational inequalities

In this paper, the existence and stability of a new system consisting of fuzzy fractional differential inclusions (FFDIs) with integral boundary conditions and variational inequalities are considered in Euclidean spaces. This system integrates the features of both FFDIs with integral boundary conditions and fractional differential variational inequalities in a unified framework. The existence of solutions for this system is established under some mild conditions. Moreover, the semicontinuity of the solution map for the perturbed system is investigated with respect to parameter perturbations. Finally, we use two numerical examples to show the validity of the above theoretical findings.

math.OC↗

arXiv2026

Agentic RAG-VLM: Affordance-Aware Retrieval-Augmented Generation with Self-Reflective Planning for Robotic Grasping

Generalizable robotic grasping in cluttered environments is essential for deploying manipulators in unstructured human spaces, yet existing VLM-based methods rely on visual similarity for object matching, neglecting physical affordances such as handle graspability and material fragility, and operate open-loop without spatial reasoning or failure recovery, limiting their effectiveness when objects are densely packed or physically diverse. We present Agentic RAG-VLM, a unified framework that bridges VLM-based semantic understanding and physically grounded grasp execution by integrating retrieval-augmented generation (RAG) with vision-language models (VLMs) and agentic self-reflective planning. Agentic RAG-VLM introduces three tightly coupled components: (1) a Hierarchical Affordance-Aware RAG (HAA-RAG) that encodes four-dimensional affordance descriptors, including type, material, fragility, and graspable region, and retrieves strategies by functional affordance compatibility rather than visual appearance; (2) a Scene Graph Constraint Reasoner that constructs spatial relationship graphs from VLM perception and translates proximity, occlusion, and support constraints into concrete grasp parameter adjustments; and (3) an Agentic Self-Reflective Pipeline with a 14-type failure taxonomy and three-level adaptive retry for closed-loop grasp refinement. Evaluated on a 12-task benchmark spanning single-grasp, interactive, and long-horizon scenarios with 360 trials per configuration, Agentic RAG-VLM achieves 78.3 percent overall success, a 53.3 percentage-point absolute gain over VLM-only baselines, demonstrating that affordance-aware retrieval, scene graph reasoning, and agentic recovery are jointly essential for robust manipulation.

cs.AI↗

arXiv2026

RoboTacDex: A Dexterous Visual-Tactile-Action Dataset for Humanoid Manipulation

In the field of robot learning, large-scale and diverse demonstration trajectories provide the fundamental basis for enhancing robotic manipulation ability. We introduce RoboTacDex, a large, multi-modal, and diverse dataset of dexterous manipulation behaviors performed with a humanoid robot. Built on the publicly accessible humanoid robot Unitree G1, RoboTacDex consists of 6k trajectories covering 19 tasks, 23 skills, and interactions with 22 objects. RoboTacDex provides comprehensive records including multi-view RGB and depth information, tactile feedback, and detailed semantic annotations. Furthermore, the dataset features a variety of relatively challenging tasks that can only be completed by dual arms and dexterous hands, aiming to mimic human-like operational logic and simulate real-world manipulation complexity. To ensure data collection quality, we develop an improved multi-camera synchronization system to enable millisecond data synchronization and recording of modalities. In our experiments, we evaluate three representative imitation learning models on our dataset, analyzing their performance as well as their respective strengths and limitations across different task categories. Successful trial results and a moderate level of generalization capabilities across a suite of tasks indicate the effectiveness and diversity of the collected dataset. Our dataset will be open-sourced soon.

cs.RO↗

arXiv2026

Trust Your Instincts: Confidence-Driven Test-Time RL for Vision-Language-Action Models

Reinforcement learning (RL) has become indispensable for pushing Vision-Language-Action Models (VLAs) beyond static imitation learning. However, existing RL methods typically require external environmental feedback, relying on predefined success signals to guide policy updates. In this work, we show that VLA models possess useful internal evaluative capabilities: in discrete-action VLAs, trajectories with higher generation confidence are significantly more likely to succeed. Based on this observation, we introduce T^2VLA (Test-time VLA), an architecture-agnostic test-time RL framework that enables VLA models to achieve self-bootstrapping policy improvement. Instead of relying on external rewards, T^2VLA leverages trajectory-level similarity to high-confidence expert demonstrations as an intrinsic reward signal. In addition, we propose a Confidence-Driven Dual Expert Bootstrapping mechanism, which dynamically balances a Local Pseudo-Expert for exploration and a Global Expert Pool for training stability. Extensive experiments on the LIBERO and RoboTwin benchmarks show that T^2VLA consistently outperforms supervised baselines and approaches oracle RL performance with ground-truth rewards, achieving effective improvement without external reward feedback. Furthermore, T^2VLA adapts to distinct VLA paradigms, including both OpenVLA-OFT and the pi series.

cs.RO↗

arXiv2026

Quadruple decomposition of boundary vorticity flux

First introduced by Lighthill in 1963 for two-dimensional flows and later generalized by Jie-Zhi Wu to three-dimensional scenarios since 1986, the boundary vorticity flux (BVF) is the cornerstone of boundary vorticity dynamics, which quantifies the vorticity source strength on a solid boundary. Recent advances in vorticity and vortex dynamics have revealed both the rigid-rotation and spin modes of vorticity from multiple perspectives. In the present study, we propose a novel quadruple decomposition of the BVF on a stationary solid wall, which essentially uncovers the boundary creation rates of the elementary vorticity modes for both the tangential and wall-normal BVF components, respectively. The proposed framework is illustrated through skin-friction and surface-pressure measurements for flow over a hill model in a low-speed wind tunnel, revealing a set of intriguing BVF patterns for the first time. These theoretical results are expected to be valuable for global surface flow diagnostics when combined with experiments, as well as for understanding the formation mechanisms of near-wall coherent structures and flow-induced noise.

physics.flu-dyn↗

arXiv2026

Towards Fast and Effective Long Video Understanding of Multimodal Large Language Models via Adaptive Quasi-Gaussian Sampling

Long video understanding remains a daunting challenge for Multimodal Large Language Models (MLLMs) due to the excessive computation and memory footprint. Thus, keyframe selection is often adopted to mitigate this shortcoming, which however still suffers from low flexibility and high noise due to its hard sampling principle. In this paper, we define video frame selection as a problem of Quasi-Gaussian Sampling, and propose an adaptive and training-free approach termed AdaQ. Inspired by the 3-$σ$ rule of Gaussian distribution, the objective of AdaQ is to achieve the optimal 3-$σ$ interval for different examples, i.e., a smaller 3-$σ$ interval for the local query and a larger one for the global query, thereby facilitating robust and adaptive frame sampling. To validate AdaQ, we apply it to four MLLMs with three embedding models. The extensive experimental results not only show its obvious performance gains over the default MLLMs and the SOTA keyframe selection methods, e.g., helping Qwen3-VL-8B outperform GPT4o by 15.8% on average by using only 64 frames, but also confirm its superior robustness and high efficiency for long-video understanding, e.g., only 1 hyper-parameter needs to be set.

cs.CV↗

arXiv2026

Towards a Dynamic and Fixed-budget Memory Bank for Efficient Streaming Video Understanding

Currently, streaming video understanding is still a daunting task for existing \emph{multimodal large language models} (MLLMs). Its difficulties not only lie in handling the ever-increasing video frames, but also in the unpredictability of future video content and input instructions. In this paper, we study this task from the perspective of constructing a dynamic but fixed-budget memory bank, and propose a novel and training-free approach termed \emph{\textbf{CausalMem}}. CausalMem is dedicated to constructing a dynamic visual memory update mechanism, thereby maximizing the amount of information in streaming video within a limited memory space, much like the human brain. In practice, CausalMem estimates the redundancy of visual tokens and updates the memory bank via an online semantic basis, which models the principal semantics of the observed video stream. To validate CausalMem, we apply it to two representative MLLMs, namely LLaVA-OneVision and Qwen2.5-VL respectively, and conduct extensive experiments on both streaming and offline video understanding benchmarks. The experimental results not only show the great advantages than existing methods under both streaming and offline settings, \emph{e.g.}, $+3.2\%$ and $+3.0\%$ average accuracy gains respectively, but also witness the superior semantic preservation for streaming videos, \emph{e.g.}, using 12$k$ token budgets to memorize hour-long streaming videos, which achieves more than \textbf{20$\times$} visual token compression ratio and only occupies about \textbf{82 MB} storage. \textbf{Our code} is given in \href{https://github.com/hktk07/CausalMem}{CausalMem}.

cs.CV↗

arXiv2026

A Robust Two-Stage Protocol for STAR-RIS-Aided ISAC Networks: Joint Beamforming and Mode Optimization

This paper investigates the robust design of integrated sensing and communication (ISAC) systems assisted by simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs), acting as programmable metasurfaces enabling concurrent sensing and communication over the full space. To exploit the dual transmission-reflection capability of STAR-RISs, we propose a two-stage ISAC protocol: a preparation phase jointly performs direction-of-arrival (DoA) estimation for outdoor users and downlink communication to all users, while a subsequent communication phase leverages the acquired angular information to enhance downlink transmission. To capture sensing uncertainty and imperfect channel knowledge, the DoAs of outdoor users are modeled as Gaussian random variables, and the non-line-of-sight (NLoS) channel components of outdoor links are characterized through their spatial covariance statistics, enabling a robust design that incorporates average communication performance into the optimization. We formulate a performance-balanced optimization problem that maximizes the communication sum-rate while guaranteeing sensing accuracy, jointly determining the beamforming vectors, the STAR-RIS transmission and reflection coefficients in both stages, and the metasurface partition between energy-splitting and transmit-only modes. To address the resulting non-convex mixed discrete-continuous problem, we develop a tailored alternating optimization framework with proven monotonic convergence. Numerical results demonstrate approximately 15% throughput gain over the most competitive benchmark neglecting NLoS statistical characterization, with robustness maintained under DoA estimation errors and imperfect NLoS channel knowledge.

eess.SP↗

arXiv2026

BCL: Bayesian In-Context Learning Framework for Information Extraction

Existing information extraction (IE) tasks increasingly adopt in-context learning (ICL) with large language models. However, current approaches either show inconsistent performance across model scales or lack systematic optimization and generalizability. Building on this, we propose BCL (Bayesian In-Context Learning Framework for Information Extraction), the first optimization framework that uses particle filtering with Bayesian updates to systematically refine label representations across IE tasks. Through four steps initialization, observation, weight update, and resampling, BCL generalizes to both sequence labeling and relation classification paradigms. Extensive experiments demonstrate substantial and consistent improvements over existing approaches.

cs.CL↗