arXiv subjects
Publications and source records attributed to Jiawen Zhang.
We develop a quantitative approach to quasi-local operators with bounded block-rank. Our main results are a quantitative decomposition for quasi-local operators into block-rank-one pieces and a quantitative finite-propagation approximation in the block-rank-one case, leading to a quantitative proof of the bounded block-rank rigidity result for Roe and quasi-local algebras.
A central question in higher index theory and operator algebras is whether the Roe algebra and the quasi-local algebra associated with a metric space of bounded geometry coincide, or at least have the same $K$-theory. In this paper, we focus on a \emph{sparse} metric space $X$. We prove the following three main results: (1) For a block-diagonal operator $T$ with uniformly bounded block-rank, $T$ is quasi-local if and only if it is in the Roe algebra. (2) In general, we discover a ghost block-diagonal projection which is quasi-local but not in the Roe algebra. (3) For a sequence of expander graphs with sufficiently large girth, the inclusion of the uniform Roe algebra into the uniform quasi-local algebra induces a \emph{non-surjective} map on their $K_0$-groups. This yields the first known $K$-theoretic distinction between the uniform Roe algebra and the uniform quasi-local algebra.
Single crystals of Ce$_5$RhGe$_2$ and Ce$_5$IrGe$_2$ have been systematically investigated by electrical resistivity, specific heat, and magnetization measurements. Together with Ce$_5$CoGe$_2$, all three compounds crystallize in the orthorhombic \emph{Pnma} structure, with the lattice parameters increasing monotonically from Co to Rh to Ir, consistent with the effect of negative chemical pressure. Magnetization measurements along the three principal crystallographic axes identify the \emph{a} axis as the easy magnetization direction throughout the series. Ce$_5$RhGe$_2$ exhibits ferromagnetic ordering with a Curie temperature of approximately 11.5 K and shows magnetic behavior closely resembling that of Ce$_5$CoGe$_2$. In contrast, Ce$_5$IrGe$_2$ undergoes two successive magnetic transitions at $T_{\rm M1}=12.7$ K and $T_{\rm M2}=11.8$ K, and there are multiple metamagnetic transitions under magnetic fields, giving rise to magnetization plateaus at fractions of the saturation magnetization $M_{\rm s}$ of approximately $M_{\rm s}/5$ and $M_{\rm s}/3$. The low-field metamagnetic transition along the easy axis shifts to lower field with decreasing temperature, and eventually a pronounced hysteresis loop is observed about zero-field, establishing that Ce$_5$IrGe$_2$ exhibits a ferrimagnetic ground state at the lowest measured temperatures.
Dislocations are being engineered into perovskite oxides to harvest versatile functional properties. One major bottleneck, however, persists: perovskite oxides that can be engineered with dislocations, particularly via mechanical deformation at room temperature in bulk scale, have so far been limited to only three materials: SrTiO3 (2001, Brunner et al.), KNbO3 (2016, Mark et al.), and KTaO3 (2024, Fang & Zhang et al.). Here, we propose a simple and effective approach by using solid solution to significantly extend the range of materials. We showcase KTa1-xNbxO3 (0<x<1) perovskite oxides for their bulk plasticity at room temperature by constructing a closed-loop validation workflow that includes crystal growth, Brinell indentation, bulk compression, and transmission electron microscopy characterization. Our findings are expected to unlock the materials toolbox for dislocation-tuned functionality of perovskite oxides.
The recently discovered heavy-fermion spin-triplet superconductor candidate UTe$_2$ provides a rich platform for unconventional pairing and topological phenomena. However, limited has been known about its superconducting order parameters and their evolution with control parameters, largely due to the lack of appropriate symmetry-sensitive detections. Here, we report comprehensive point-contact spectroscopy measurements of pressurized UTe$_2$ on the (0~0~1) surface. The observation of Andreev bound states strongly suggests the presence of a $p_z$ component in the superconducting order parameters. Quantitative analysis based on an extended Blonder-Tinkham-Klapwijk model unveils the superconducting order parameters with a finite odd-$k_z$ component (e.g. $B_{2u}$ or $B_{3u}$) for both ambient and pressurized UTe$_2$. Remarkably, the multiple superconducting phases can be distinguished by a single parameter $\langle Δ_{z}\rangle/\langleΔ_{x(y)}\rangle$, the relative weight between the $p_z$-wave and $p_{x(y)}$-wave pairings. These findings place stringent constraints on the pairing symmetry and provide essential spectroscopic signatures for distinguishing pressure-induced multiple superconducting phases in UTe$_2$.
When a user question is underspecified, a capable model should recognize that its context is insufficient, identify the missing information, ask for it, and respond only once that information determines a unique answer. We formalize multi-turn information seeking as solving a k-underspecified constraint satisfaction problem, where k is the number of variables jointly required to determine the target and therefore measures the degree of missing information. We instantiate the formulation in MT-InfoSeek, a controlled evaluation suite of 5,251 problems and 9,006 task instances spanning mathematics, logic, biology, medicine, and general knowledge. We evaluate models along three axes: what they ask, when they ask it, and how the acquired information affects the final answer. Performance degrades across models and domains as underspecification increases. Models recognize that additional information is needed but underestimate how much, and in logical problems at k = 2 they under-predict the degree of missing information about four times as often as they over-predict it. They also fail to identify a minimal sufficient set of queries, improve only marginally when given the true k, and often stop before acquiring sufficient information. In tasks with ordered dependencies, an incorrect query order reduces final accuracy even when the model eventually acquires all necessary information. We measure information seeking directly through final sufficiency, which records whether the acquired information determines the target independent of answer generation. This separation shows differences between models that final accuracy alone does not capture, and indicates that the ability to seek information over multiple turns is distinct from the ability to generate answers and is not measured by current LLM evaluations.
Time series reasoning tasks often start with a natural language question and require targeted analysis of a time series. Evidence may span the full series or appear in a few short intervals, so the model must decide what to inspect. Most existing approaches encode the entire time series into a fixed representation before inference, regardless of whether or not the entire sequence is relevant. We introduce ARTIST, which formulates time-series reasoning as a sequential decision problem. ARTIST interleaves reasoning with adaptive temporal segment selection. It adopts a controller-reasoner architecture and uses reinforcement learning to train the controller role to select informative segments and the reasoner role to generate segment-conditioned reasoning traces and final answers. During inference, the model actively acquires task-relevant information instead of relying on a static summary of the full sequence. We use a novel hierarchical policy optimization approach for post-training that allows the model to excel in both segment selection and question-answering behavior. We evaluate ARTIST on six time-series reasoning benchmarks and compare it with large language models, vision-language models, and prior time-series reasoning systems. ARTIST improves average accuracy by 6.46 absolute percentage points over the strongest baseline. The largest gains appear on rare event localization and multi-segment reasoning tasks. Supervised fine-tuning improves performance, and reinforcement learning provides additional gains by optimizing question-adaptive segment selection. These results show that selective data use drives effective time-series reasoning.
The kagome metal CsV$_{3}$Sb$_{5}$ hosts an intriguing interplay between charge-density-wave (CDW) order and superconductivity that is highly sensitive to lattice distortions. However, determining the specific roles of the in-plane ($A_{1g,1}$) and out-of-plane ($A_{1g,2}$) symmetric strain channels has been hindered by their intrinsic mixing in conventional piezo-based experiments. Here, we combine in-plane uniaxial strain with direct $c$-axis compression to independently access and disentangle these symmetry-resolved responses in CsV$_{3}$Sb$_{5}$. We reveal that $c$-axis compression drives a massive, linear enhancement of the superconducting transition temperature ($T_c$) alongside a suppression of $T_{\rm CDW}$. The tuning efficiency of this out-of-plane deformation acts with an opposite sign and far exceeds that of in-plane strain, demonstrating that $c$-axis lattice control dictates the phase competition. Furthermore, by isolating the pure elastoresistivity coefficients, we find that the out-of-plane cross-coupling coefficient ($m_{13}$) is comparable in magnitude but opposite in sign to the in-plane response ($m_{11}+m_{12}$). Unlike the sharply peaked in-plane response, $m_{13}$ exhibits a distinct, order-parameter-like onset across the CDW transition. Our results establish that out-of-plane lattice control plays a dominant role in tuning the intertwined states in CsV$_{3}$Sb$_{5}$ and provide a general pathway for resolving strain-coupled electronic responses in layered quantum materials.
Wireless aerial virtual reality (VR) aims to provide immersive access to large-scale scenes, but high-resolution view generation and delivery are jointly constrained by limited bandwidth, latency, and power. 3D Gaussian Splatting (3DGS) can reduce the payload by rendering views from compact pose information, yet its geometry errors may cause severe VR quality degradation. Existing channel-aware or pixel-level resource allocation schemes fail to capture such geometry-sensitive distortion. To address this issue, this paper proposes GeoFovea-GS as a geometry-aware cross-layer framework for communication-efficient wireless aerial VR. A foveated geometry-aware distortion metric is developed to characterize photometric rendering error, geometric inconsistency, and view-dependent perceptual importance in a unified form. Based on this metric, the joint selection of pose-only 3DGS rendering and image/tile correction transmission is formulated as a cross-layer optimization problem under wireless constraints. A lightweight value-of-information scheduler is further developed to allocate communication resources to regions that are both geometry-critical and perceptually important. Experiments on real-world 3DGS scenes demonstrate that GeoFovea-GS achieves superior immersive rendering quality with substantially reduced transmission cost.
The study of vacuum is important in understanding compressible flows. In particular, physical vacuum, in which the boundary moves with a nontrivial finite normal acceleration, naturally arises in the study of the motion of gaseous stars. In this paper, we analyze the free boundary problem for the three-dimensional compressible Navier--Stokes--Poisson equations with degenerate viscosities for self-gravitating viscous gaseous stars. For the spherically symmetric and barotropic motion, we establish the global well-posedness of classical solutions without any restriction on the size of the initial data. Our solutions obtained here are smooth all the way up to the moving boundary and capture the physical vacuum boundary behavior of the Lane--Emden star configuration.
Persistent personal agents combine long-term memory with access to users' external environments, enabling personalized foreground assistance and proactive background execution. This integration also creates a new path to compromise: untrusted external content can be silently written into persistent memory and later reused as trusted state. We study this threat as stealth memory injection, in which a remote black-box adversary delivers a single email payload that must induce the agent to write poisoned memory, stay hidden in the agent's response to the user, and affect future behavior. We introduce WhisperBench, a 108-case benchmark spanning five risk categories and both fact and preference poisoning. Built on a real IMAP/SMTP workflow and an authentic email agent skill, it enables full-cycle evaluation of stealth memory injection attacks. To enable this black-box attack under single-email delivery and without runtime feedback, we propose MemGhost, a one-shot payload generation framework. MemGhost uses an environment proxy to emulate persistent-agent execution and an objective proxy to convert memory adoption and conversational stealth into dense rubric-based rewards, then trains the attacker policy with supervised fine-tuning and reinforcement learning. Across 56 held-out test cases, MemGhost achieves 87.5% end-to-end success on OpenClaw with GPT-5.4 and 71.4% on Claude Code SDK with Sonnet 4.6. It also transfers across personal-agent architectures (NanoClaw and Hermes Agent) and memory backends (filesystem and vector-based Mem0), and remains effective against input-level, model-level, and system-level defenses. These results suggest that persistent memory can turn ordinary external processing into a practical pathway for long-term agent compromise.
This paper focuses on the free boundary problem of the three-dimensional compressible Navier-Stokes-Poisson equations with degenerate viscosities for self-gravitating viscous gaseous stars. For spherically symmetric barotropic motion, we establish the local well-posedness of classical solutions. The solutions obtained here are smooth all the way up to the moving boundary and capture the physical vacuum boundary behavior of the Lane-Emden star configuration for all adiabatic exponents $γ>\frac{4}{3}$.
Indirect prompt injection attacks hijack LLM-based agents by embedding malicious instructions in third-party data that the agent retrieves during task execution. Existing defenses report near-zero attack success rate on static benchmarks, yet recent adaptive evaluations show that these results collapse once the attacker is allowed to optimize against the deployed defense. In this work, we trace this collapse to two failure modes. First, existing defense methods are confined to recognizing specific attack patterns, rather than assessing whether the intent of every embedded instruction is relevant to the user task. Second, training-based defenses, which otherwise offer the strongest safety-utility trade-off, assemble their adversarial examples from a handful of hand-crafted templates, and the resulting defender fails to generalize outside that narrow strategy distribution. To address these gaps, we propose RETA, a training-based method that grounds defense decisions on the user tasks rather than attacker-controlled data. At each tool-output step, the defender undertakes chain-of-thought reasoning verifying that its actions are consistent with the user task. Leveraging red-teaming, a simulated attacker synthesizes adversarial training data and receives a dictionary-learning diversity reward, achieving broad coverage of injection-reformulation strategies. Together, these allow the defender to be optimized via multi-objective reinforcement learning and achieve better safety-utility trade-off. Across six black-box adaptive attacks, RETA keeps every per-attack ASR below 10%, with average ASR of 2.92% and 3.75% on the two target models, while preserving most utility under attack and on clean inputs.
Evaporating sessile droplets are usually treated as capillary objects, but droplets generated by routine handling can carry tens to hundreds of picocoulombs of electric charge. Here we combine Faraday-cup charge measurements with optical imaging to determine how such charge evolves as water droplets evaporate on dry polymer substrates. A zero-time protocol shows that a reproducible initial charge is preserved on poly(methylpentene) (PMP), whereas PDMS, SOCAL-coated surfaces, and polystyrene either exchange, dissipate, or inject charge on contact. On PMP, ensemble-resolved measurements reveal two regimes: the charge remains nearly constant during early evaporation and then decreases abruptly once the droplet reaches a small-volume state. This charge collapse coincides with spontaneous lateral translation rather than jetting or breakup. A Rayleigh-normalized analysis, including a spherical-cap stress correction and measured contact-angle retention scale, shows that motion occurs only after evaporation drives the droplet into a high electro-pinning state. High-speed imaging and kinematic analysis support a picture in which the subsequent motion is governed by repeated contact-line depinning and re-pinning: the total distance traveled is strongly affected by dry-surface pinning, whereas the peak translational velocity serves as a more robust indicator of the discharge strength. These results identify a dry-substrate mode of evaporation-driven electrostatic relaxation, distinct from Coulomb fission on lubricated surfaces, in which substrate electrostatic passivity enables charge retention, droplet geometry selects the instability onset, and whole-droplet translation provides the charge-release pathway.
Personal AI agents increasingly rely on long-term memory to provide persistent personalization across sessions. However, existing memory pipelines are largely driven by semantic similarity: memory data close to the current query is retrieved and injected into the model context. This creates a critical trustworthiness gap, since a semantically related memory may still be contextually inappropriate, leading to threats such as cross-domain leakage, sycophancy, tool-call drift, or memory-induced jailbreaks. In this paper, we study memory search as a trust boundary in personal AI agents. We evaluate representative agentic memory frameworks, including A-Mem, Mem0, and MemOS, together with OpenClaw, a real-world personal-agent environment with persistent state and tool-use capability. Our results show that long-term memory is not merely a utility layer, but a durable control channel that can reshape how agents interpret tasks and execute actions, leaving them highly susceptible to the aforementioned threats. To mitigate these vulnerabilities, we propose MemGate, a lightweight and deployable memory plug-in for trustworthy memory search, with only 9M parameters and a 35.1MB footprint. MemGate is inserted between the vector memory store and the backbone LLM, requiring no LLM modification, memory-database rewriting, or inference-time LLM judge. It applies a query-conditioned neural gate to candidate memory representations, turning raw similarity search into task-conditioned memory admission. Across multiple mainstream memory frameworks, real-world agent settings, and diverse LLM backbones, MemGate reduces memory-induced threats while preserving long-term memory utility.
While Large Language Models (LLMs) demonstrate remarkable capabilities, they remain susceptible to sophisticated, multi-step jailbreak attacks that circumvent conventional surface-level safety alignment by exploiting the internal generation process. To address these vulnerabilities, we propose Reflector, a principled two-stage framework that internalizes self-reflection within the generation trajectory. Reflector first leverages teacher-guided generation to produce high-quality reflection data for supervised fine-tuning (SFT), establishing structured reflection patterns. It subsequently uses Reinforcement Learning (RL) with outcome-driven and reward-validity supervision to instill robust, autonomous self-reflection capabilities. Empirical results show that Reflector achieves Defense Success Rates (DSR) exceeding 90% against complex indirect attacks while generalizing robustly across diverse threat scenarios. Notably, the framework enhances both task-specific and general utility, yielding a 5.85% gain on GSM8K alongside improved performance on knowledge-intensive benchmarks. By internalizing trajectory-level safety, Reflector overcomes the fundamental limitations of surface alignment without significant computational overhead, offering an efficient and scalable solution for the development of safe and capable LLMs.
Long-form video generation is rapidly moving from short, single-scene synthesis toward minute-long, multi-shot creation with narrative structure, cinematic control, audio, and cross-modal synchronization. However, evaluating such videos remains challenging, since existing benchmarks largely focus on local visual quality, short-horizon temporal consistency, or generic prompt alignment, and provide limited diagnosis of workflow failures and user-dependent preferences. We introduce DirectorBench, a personalized multi-agent diagnostic benchmark for long-form video generation. DirectorBench evaluates generated videos with respect to 80 structured metadata entries, 7 user profiles, and 40 checkpoint criteria across 5 dimensions: script, visual, audio, cross-modal, and stability. Instead of reducing quality to a single aggregate score, DirectorBench localizes checkpoint-level bottlenecks and supports profile-aware evaluation. We evaluate 4 long-form video generation workflows, 6 base LLMs, and 7 user profiles. Across workflows, DirectorBench reveals a between-unit bottleneck: transition quality averages only 0.256 and reaches 0.356 for the best workflow, while prompt-level user demand fulfillment averages 0.71. We further conduct human evaluation with 14 annotators to validate the alignment between DirectorBench and human judgment. The results show that DirectorBench captures human-perceptible quality differences and reveals workflow- and profile-dependent failure modes that are hidden by aggregate scoring. These findings highlight the importance of diagnostic and profile-aware benchmarking for long-form video generation.
While 3D generation is progressing rapidly, recent work has often focused on obtaining high-resolution assets, leaving user experience and deployability as afterthoughts. We present AssetGen, a 3D generator that focuses instead on these two aspects. Given one reference image, in 30 seconds it produces a high-quality mesh with baked normals, a color texture, and a controlled polygon budget suitable for real-time rendering, including mobile use cases. The AssetGen Flash variant further reduces latency to 14 seconds for interactive and agentic creation loops. Our model generates the object geometry with a coarse-to-refine VecSet framework, which implements mesh simplification, cleaning, and normal baking on the GPU, and a fast parallel UV unwrapping. It then generates textures in a multi-view fashion, followed by backprojection and 3D inpainting. Model distillation, kernel optimization, and pipeline parallelization are co-designed to accelerate the system end-to-end. We introduce numerous automated and blind human evaluations and demonstrate competitive visual quality against leading commercial solutions in 30 seconds and preview-quality results in less than 15 seconds. The final result is a system that supports AI-assisted, deployable 3D content creation in interactive workflows.