arXiv subjects

Teng Ma

Publications and source records attributed to Teng Ma.

At least 37 records · Page 2Linked to original sources

arXiv2026

SiriusHelper: An LLM Agent-Based Operations Assistant for Big Data Platforms

Big data platforms are widely used in modern enterprises, and an in-production intelligent assistant is increasingly important to help users quickly find actionable guidance and reduce operational burden. While recent LLM+RAG assistants provide a natural interface, they face practical challenges in real deployments: limited scenario coverage across both general consultation and domain-specific troubleshooting workflows, inefficient knowledge access due to inadequate multi-hop retrieval and flat knowledge organization, and high maintenance cost because escalated tickets are unstructured and hard to convert into assistant improvements and reusable SOPs. In this paper, we present SiriusHelper, a deployed intelligent assistant for big data platforms. SiriusHelper serves as a unified online assistant that automatically identifies user intent and routes queries to the right handling path, including dedicated expert workflows for specialized scenarios (e.g., SQL execution diagnosis). To support complex troubleshooting, SiriusHelper combines a DeepSearch-driven mechanism with a priority-based hierarchical knowledge base to enable multi-hop retrieval without context overload, thus improving answer reliability and latency. To reduce expert overhead, SiriusHelper further introduces automated ticket understanding and SOP distillation: it diagnoses the assistant failure reason (e.g., missing knowledge or wrong routing) and extracts domain-specific SOPs to continuously enrich the knowledge base. Experiments and online deployment on Tencent Big Data platform show that SiriusHelper outperforms representative alternatives and reduces online ticket volume by 20.8\%.

cs.DB↗

arXiv2026

UniSonate: A Unified Model for Speech, Music, and Sound Effect Generation with Text Instructions

Generative audio modeling has largely been fragmented into specialized tasks, text-to-speech (TTS), text-to-music (TTM), and text-to-audio (TTA), each operating under heterogeneous control paradigms. Unifying these modalities remains a fundamental challenge due to the intrinsic dissonance between structured semantic representations (speech/music) and unstructured acoustic textures (sound effects). In this paper, we introduce UniSonate, a unified flow-matching framework capable of synthesizing speech, music, and sound effects through a standardized, reference-free natural language instruction interface. To reconcile structural disparities, we propose a novel dynamic token injection mechanism that projects unstructured environmental sounds into a structured temporal latent space, enabling precise duration control within a phoneme-driven Multimodal Diffusion Transformer (MM-DiT). Coupled with a multi-stage curriculum learning strategy, this approach effectively mitigates cross-modal optimization conflicts. Extensive experiments demonstrate that UniSonate achieves state-of-the-art performance in instruction-based TTS (WER 1.47%) and TTM (SongEval Coherence 3.18), while maintaining competitive fidelity in TTA. Crucially, we observe positive transfer, where joint training on diverse audio data significantly enhances structural coherence and prosodic expressiveness compared to single-task baselines. Audio samples are available at https://qiangchunyu.github.io/UniSonate/.

eess.AS↗

arXiv2026

One-shot learning for the complex dynamical behaviors of weakly nonlinear forced oscillators

Extrapolative prediction of complex nonlinear dynamics remains a central challenge in engineering. This study proposes a one-shot learning method to identify global frequency-response curves from a single excitation time history by learning governing equations. We introduce MEv-SINDy (Multi-frequency Evolutionary Sparse Identification of Nonlinear Dynamics) to infer the governing equations of non-autonomous and multi-frequency systems. The methodology leverages the Generalized Harmonic Balance (GHB) method to decompose complex forced responses into a set of slow-varying evolution equations. We validated the capabilities of MEv-SINDy on two critical Micro-Electro-Mechanical Systems (MEMS). These applications include a nonlinear beam resonator and a MEMS micromirror. Our results show that the model trained on a single point accurately predicts softening/hardening effects and jump phenomena across a wide range of excitation levels. This approach significantly reduces the data acquisition burden for the characterization and design of nonlinear microsystems.

cs.LG↗

arXiv2026

Genesis of a Horizontal Electric Field within the Lipid Bilayer: A Bilayer-Embedded Actuation Platform

The electric field of biological membranes has long been treated as a one-dimensional quantity, defined solely by the component normal to the bilayer (E_VERT). Here, we present a bioelectronic platform that enables controlled generation of a horizontal electric field within the hydrophobic core of a lipid bilayer (E_HORZ). The device incorporates micrometer-scale electrodes embedded within the bilayer torus, allowing sustained E_HORZ actuation. Applied E_HORZ selectively and reversibly accelerates slow inactivation of a voltage-gated potassium channel while leaving activation essentially unchanged. Physical considerations further indicate that E_HORZ arises naturally wherever membrane potential varies spatially, including at action potential wavefronts, suggesting broader physiological relevance. This platform provides experimental access to vector-resolved membrane electric fields and establishes a generalizable strategy for multidirectional electrical control of soft-matter biointerfaces.

physics.bio-ph↗

arXiv2026

V-CAST: Video Curvature-Aware Spatio-Temporal Pruning for Efficient Video Large Language Models

Video large language models (VideoLLMs) show strong capability in video understanding, yet long-context inference is still dominated by massive redundant visual tokens in the prefill stage. We revisit token compression for VideoLLMs under a tight budget and identify a key bottleneck, namely insufficient spatio-temporal information coverage. Existing methods often introduce discontinuous coverage through coarse per-frame allocation or scene segmentation, and token merging can further misalign spatio-temporal coordinates under MRoPE-style discrete (t,h,w) bindings. To address these issues, we propose V-CAST (Video Curvature-Aware Spatio-Temporal Pruning), a training-free, plug-and-play pruning policy for long-context video inference. V-CAST casts token compression as a trajectory approximation problem and introduces a curvature-guided temporal allocation module that routes per-frame token budgets to semantic turns and event boundaries. It further adopts a dual-anchor spatial selection mechanism that preserves high-entropy visual evidence without attention intervention, while keeping retained tokens at their original coordinates to maintain positional alignment. Extensive experiments across multiple VideoLLMs of different architectures and scales demonstrate that V-CAST achieves 98.6% of the original performance, outperforms the second-best method by +1.1% on average, and reduces peak memory and total latency to 86.7% and 86.4% of vanilla Qwen3-VL-8B-Instruct.

cs.CV↗

arXiv2026

TrEnv-X: Transparently Share Serverless Execution Environments Across Different Functions and Nodes

Serverless computing is renowned for its computation elasticity, yet its full potential is often constrained by the requirement for functions to operate within local and dedicated background environments, resulting in limited memory elasticity. To address this limitation, this paper introduces TrEnv-X, a co-designed integration of the serverless platform with the operating system and CXL/RDMA-based remote memory pools. TrEnv-X's core innovations are repurposable sandboxes, which can be shared across different functions to decrease the associated creation overhead, and OS-level memory templates, which enable rapid state restoration from CXL/RDMA-based remote memory pools. To further demonstrate TrEnv-X's versatility, we generalize its design from traditional containers for microVM-based agent workloads and introduce new optimizations, including browser sharing and a page cache bypassing mechanism. Our evaluation shows that TrEnv-X achieves up to 7x reduction in P99 latency and 48% memory savings for container-based functions. When applied to LLM agents, it reduces the P99 latency by up to 58% and memory usage by 61% compared to state-of-the-art systems like E2B.

cs.DC↗

arXiv2026

EcoAlign: An Economically Rational Framework for Efficient LVLM Alignment

Large Vision-Language Models (LVLMs) exhibit powerful reasoning capabilities but suffer sophisticated jailbreak vulnerabilities. Fundamentally, aligning LVLMs is not just a safety challenge but a problem of economic efficiency. Current alignment methods struggle with the trade-off between safety, utility, and operational costs. Critically, a focus solely on final outputs (process-blindness) wastes significant computational budget on unsafe deliberation. This flaw allows harmful reasoning to be disguised with benign justifications, thereby circumventing simple additive safety scores. To address this, we propose EcoAlign, an inference-time framework that reframes alignment as an economically rational search by treating the LVLM as a boundedly rational agent. EcoAlign incrementally expands a thought graph and scores actions using a forward-looking function (analogous to net present value) that dynamically weighs expected safety, utility, and cost against the remaining budget. To prevent deception, path safety is enforced via the weakest-link principle. Extensive experiments across 3 closed-source and 2 open-source models on 6 datasets show that EcoAlign matches or surpasses state-of-the-art safety and utility at a lower computational cost, thereby offering a principled, economical pathway to robust LVLM alignment.

cs.AI↗

arXiv2026

Pooling Engram Conditional Memory in Large Language Models using CXL

Engram conditional memory has emerged as a promising component for LLMs by decoupling static knowledge lookup from dynamic computation. Since Engram exhibits sparse access patterns and supports prefetching, its massive embedding tables are well-suited for offloading to lower-tier memory. In this paper, we propose using Compute Express Link (CXL) memory pool for Engram storage. Compared to RDMA, CXL provides fine-grained and low-latency access required by minimal and discrete retrieval patterns of Engram. We integrate the CXL-based Engram pool into SGLang, achieving near-DRAM end-to-end performance. This provides a scalable and cost-efficient storage solution for future Engram-integrated LLMs without compromising inference performance.

cs.AR↗

arXiv2026

Adaptive Multi-Objective Tiered Storage Configuration for KV Cache in LLM Service

The memory-for-computation paradigm of KV caching is essential for accelerating large language model (LLM) inference service, but limited GPU high-bandwidth memory (HBM) capacity motivates offloading the KV cache to cheaper external storage tiers. While this expands capacity, it introduces the challenge of dynamically managing heterogeneous storage resources to balance cost, throughput, and latency under varying workloads. We formulate this as a multi-objective optimization problem: identifying the Pareto frontier across these metrics within the storage configuration space. Using a high-fidelity end-to-end simulator, we observe that the objective functions are non-analytic and exhibit complex variable coupling, making the Pareto frontier difficult to approximate analytically. To obtain the frontier, we introduce Kareto, a KV-cache Adaptive REsource managemenT Optimizer. Kareto leverages a diminishing-return-guided pruning method to efficiently navigate the large configuration space and approximate the Pareto frontier. Additionally, it incorporates a fine-grained adaptive tuner that uses eviction policies in tier storage and KV block access patterns for group-specific cache management, improving cache efficiency. Experiments on real-world traces show that Kareto adapts to workload and can identify configurations of better cost efficiency, covering static strategies. Compared to the fixed setup with 1024 GB DRAM, Kareto can improve throughput by up to 9.3%, or reduce latency by up to 58.3%, or lower cost by up to 20.2% under respective optimization objectives.

cs.AR↗

arXiv2025

OmniSafeBench-MM: A Unified Benchmark and Toolbox for Multimodal Jailbreak Attack-Defense Evaluation

Recent advances in multi-modal large language models (MLLMs) have enabled unified perception-reasoning capabilities, yet these systems remain highly vulnerable to jailbreak attacks that bypass safety alignment and induce harmful behaviors. Existing benchmarks such as JailBreakV-28K, MM-SafetyBench, and HADES provide valuable insights into multi-modal vulnerabilities, but they typically focus on limited attack scenarios, lack standardized defense evaluation, and offer no unified, reproducible toolbox. To address these gaps, we introduce OmniSafeBench-MM, which is a comprehensive toolbox for multi-modal jailbreak attack-defense evaluation. OmniSafeBench-MM integrates 13 representative attack methods, 15 defense strategies, and a diverse dataset spanning 9 major risk domains and 50 fine-grained categories, structured across consultative, imperative, and declarative inquiry types to reflect realistic user intentions. Beyond data coverage, it establishes a three-dimensional evaluation protocol measuring (1) harmfulness, distinguished by a granular, multi-level scale ranging from low-impact individual harm to catastrophic societal threats, (2) intent alignment between responses and queries, and (3) response detail level, enabling nuanced safety-utility analysis. We conduct extensive experiments on 10 open-source and 8 closed-source MLLMs to reveal their vulnerability to multi-modal jailbreak. By unifying data, methodology, and evaluation into an open-source, reproducible platform, OmniSafeBench-MM provides a standardized foundation for future research. The code is released at https://github.com/jiaxiaojunQAQ/OmniSafeBench-MM.

cs.CR↗

arXiv2025

Heavy Long-lived Dark Vector Via a Gluonic Portal

We study a dark gauge boson $ Z' $ that exclusively couples to the QCD gluons through higher dimensional operators. These operators are generated from integrating out of heavy ultraviolet resonances carrying both QCD and dark gauge charges. With $ SU(3)_C $ gauge invariance, charge and parity symmetries preserved, we find that the leading effective operators are restricted to have the form of $ Z'GGG $ and $ Z'Z'GG $ at dimension-eight, which can naturally render the $Z^\prime$ particle long-lived, and serve as a viable dark matter candidate. We investigate the phenomenology of these operators with both collider experiments and cosmological observation, without and with the assumption that this dark gauge boson plays the role of the dominant dark matter component. For an unstable $Z'$, we show that depending on its lifetime, it can be probed by various observables up to ultraviolet physics scale around $10^9$ GeV. For $Z'$ being dark matter, we find that $m_{Z^\prime} \gtrsim 1 $ TeV is consistent with the thermal freeze-out scenario. In contrast, in the freeze-in scenario, the extremely small couplings leave the relevant parameter space largely unconstrained by current experiments.

hep-ph↗

arXiv2025

Quantifying EFT Uncertainties in LHC Searches

Effective Field Theory (EFT) is a general framework to parametrize the low-energy approximation to a UV model that is widely used in model-independent searches for new physics. The use of EFTs at the LHC can suffer from a 'validity' issue, since new physics amplitudes often grow with energy and the kinematic regions with the most sensitivity to new physics have the largest theoretical uncertainties. We propose a method to account for these uncertainties with the aim of producing robust model-independent results with a well-defined statistical interpretation. In this approach, one must specify the new operators being studied as well as the new physics cutoff $M$, the energy scale where the EFT approximation breaks down. At energies below $M$, the EFT uncertainties are accounted for by adding additional higher dimensional operators with coefficients that are treated as nuisance parameters. The size of the nuisances are governed by a prior likelihood function that incorporates information about dimensional analysis, naturalness, and the scale $M$. At energies above $M$, our method incorporates the lack of predictivity of the EFT, and we show that this is crucial to obtain consistent results. We perform a number of tests of this method in a simple toy model, illustrating its performance in analyses aimed at new physics exclusion as well as for discovery. The method is conveniently implemented by the technique of event reweighting and is easily ported to realistic LHC analyses. We find that the procedure converges quickly with the number of nuisance parameters and is conservative when compared to UV models. The paper gives a precise meaning and offers a principled and practical solution to the widely debated 'EFT validity issue'.

hep-ph↗

arXiv2025

TokenSim: Enabling Hardware and Software Exploration for Large Language Model Inference Systems

The increasing demand for large language model (LLM) serving has necessitated significant advancements in the optimization and profiling of LLM inference systems. As these models become integral to a wide range of applications, the need for efficient and scalable serving solutions has grown exponentially. This work introduces TokenSim, a comprehensive hardware and software exploration system designed specifically for LLM inference. TokenSim is characterized by its support for extensible system optimizations including scheduling and memory management. We validate the results with systems running with realworld datasets, achieving an error rate of less than 1%. Furthermore, TokenSim facilitates various insightful explorations into the performance and optimization of LLM serving systems.

cs.DC↗

arXiv2025

An Inorganic Liquid Crystalline Dispersion with 2D Ferroelectric Moieties

Electro-optical effect based liquid crystal devices have been extensively used in optical modulation techniques, in which the Kerr coefficient reflects the sensitivity of the liquid crystals and determines the strength of the device operational electric field. The Peterlin-Stuart theory and the O'Konski model jointly indicate that a giant Kerr coefficient could be obtained in a material with both a large geometrical anisotropy and an intrinsic polarization, but such a material is not yet reported. Here we reveal a ferroelectric effect in a monolayer two-dimensional mineral vermiculite. A large geometrical anisotropy factor and a large inherent electric dipole together raise the record value of Kerr coefficient by an order of magnitude, till $3.0\times 10^{-4}$ m V$^{-2}$. This finding enables an ultra-low operational electric field of $10^2$-$10^4$ V m$^{-1}$ and the fabrication of electro-optical devices with an inch-level electrode separation, which is not practical previously. Because of its high ultraviolet stability (decay <1% under ultraviolet exposure of 1000 hours), large-scale, and energy-efficiency, prototypical displayable billboards have been fabricated for outdoor interactive scenes. The work provides new insights for both liquid crystal optics and two-dimensional ferroelectrics.

cond-mat.mtrl-sci↗

arXiv2025

Bootstrapping the Chiral-Gravitational Anomaly

We analyze causality and unitarity constraints in graviton scattering amplitudes, aiming to establish new bounds on theories with $U(1)$-gravitational anomalies, such as axion models or strongly-coupled gauge theories. For this purpose, we show the necessity of coupling these theories to gravity. We obtain a universal scale $Λ_{\rm caus}$ at which states with $J\geq 4$ must appear in the theory. We show that this scale can lie below the quantum gravity scale. For axion models, we get $Λ_{\rm caus}\sim\sqrt{M_P f_a}$ where $f_a$ is the axion decay constant. In strongly-coupled gauge theories in the large-$N_c$ limit, the presence of glueballs allows to evade these bounds, provided the number of fermions $N_F\ll N_c$ and the 'tHooft coupling is not large. Nevertheless, for models that have a holographic 5D dual (large 'tHooft coupling), $Λ_{\rm caus}$ emerges as a new cutoff scale, unless certain conditions on the parameters of the 5D models are satisfied.

hep-th↗

arXiv2025

Hardware implementation of timely reliable Bayesian decision-making using memristors

Brains perform decision-making by Bayes theorem. The theorem quantifies events as probabilities and, based on probability rules, renders the decisions. Learning from this, Bayes theorem can be applied to enable efficient user-scene interactions. However, given the probabilistic nature, implementing Bayes theorem in hardware using conventional deterministic computing can incur excessive computational cost and decision latency. Though challenging, here we present a probabilistic computing approach based on memristors to implement the Bayes theorem. We integrate memristors with Boolean logics and, by exploiting the volatile stochastic switching of the memristors, realise probabilistic logic operations, key for hardware Bayes theorem implementation. To empirically validate the efficacy of the hardware Bayes theorem in user-scene interactions, we develop lightweight Bayesian inference and fusion hardware operators using the probabilistic logics and apply the operators in road scene parsing for self-driving, including route planning and obstacle detection. The results show our operators can achieve reliable decisions in less than 0.4 ms (or equivalently 2,500 fps), outperforming human decision-making and the existing driving assistance systems.

cs.LG↗

arXiv2025

Mining Platoon Patterns from Traffic Videos

Discovering co-movement patterns from urban-scale video data sources has emerged as an attractive topic. This task aims to identify groups of objects that travel together along a common route, which offers effective support for government agencies in enhancing smart city management. However, the previous work has made a strong assumption on the accuracy of recovered trajectories from videos and their co-movement pattern definition requires the group of objects to appear across consecutive cameras along the common route. In practice, this often leads to missing patterns if a vehicle is not correctly identified from a certain camera due to object occlusion or vehicle mis-matching. To address this challenge, we propose a relaxed definition of co-movement patterns from video data, which removes the consecutiveness requirement in the common route and accommodates a certain number of missing captured cameras for objects within the group. Moreover, a novel enumeration framework called MaxGrowth is developed to efficiently retrieve the relaxed patterns. Unlike previous filter-and-refine frameworks comprising both candidate enumeration and subsequent candidate verification procedures, MaxGrowth incurs no verification cost for the candidate patterns. It treats the co-movement pattern as an equivalent sequence of clusters, enumerating candidates with increasing sequence length while avoiding the generation of any false positives. Additionally, we also propose two effective pruning rules to efficiently filter the non-maximal patterns. Extensive experiments are conducted to validate the efficiency of MaxGrowth and the quality of its generated co-movement patterns. Our MaxGrowth runs up to two orders of magnitude faster than the baseline algorithm. It also demonstrates high accuracy in real video dataset when the trajectory recovery algorithm is not perfect.

cs.CV↗

arXiv2024

Anomalous Heat Transfer in Nonequilibrium Quantum Systems

Anomalous heat transfer (AHT), a process by which heat spontaneously flows from a cold system into a hot one, superficially contradicts the Clausius statement of the second law of thermodynamics. Here we provide a full classification of mechanisms of the AHT in nonequilibrium quantum systems from a quantum-information perspective. For initial states in local equilibrium, we find that the AHT can arise from three resources: initial correlation, intrasystem interaction, and intrasystem temperature inhomogeneity. In particular, for qubit systems, we prove that initial quantum coherence is necessary for AHT if the intersystem interactions are limited to the two-body type. We explicitly show the AHT dominated by each of the mechanisms in a three-qubit system. Our classification scheme may offer a guideline for developing high-efficiency quantum autonomous thermal machines.

quant-ph↗