Search arXiv⌕ Search

arXiv subjects

Hao Zheng

Publications and source records attributed to Hao Zheng.

At least 19 recordsLinked to original sources
arXiv2026

MultiFair: Multimodal Balanced Fairness-Aware Medical Classification with Dual-Level Gradient Modulation

Medical decision systems increasingly rely on data from multiple sources to ensure reliable and unbiased diagnosis. However, existing multimodal learning models fail to achieve this goal because they often overlook two critical challenges. First, various data modalities may learn unevenly, thereby converging to a model biased towards certain modalities. Second, the model may emphasize learning on certain demographic groups causing unfair performances. The two aspects can influence each other, as different data modalities may favor respective groups during optimization, leading to both imbalanced and unfair multimodal learning. This paper proposes a novel approach called MultiFair for multimodal medical classification, which addresses these challenges with a dual-level gradient modulation process. MultiFair dynamically modulates training gradients regarding the optimization direction and magnitude at both data modality and group levels. We evaluate MultiFair on three real-world medical classification datasets with diverse demographic attributes,including multiclass classification and missing-modality settings. Experimental results demonstrate its effectiveness.

cs.LG↗
arXiv2026

A State Augmented Approach to Paralle Modeling in High-Level Synthesis

High-level synthesis (HLS) can schedule only the concurrency made legal by its input description. For feedback-coupled designs, a sequential behavioral model can obscure the architectural state boundaries and hazard resolution needed for efficient overlap. We develop an explicit architectural modeling methodology based on synchronous sequential processes that start in lockstep and synchronize completion before repeating. Persistent state, communication, and feedback control are represented in C++, while HLS schedules and binds operations within that structure. A five-stage RV32I processor illustrates three consequential modeling decisions: which version of a value a consumer reads, when it consumes that value, and how architectural memory actions map to port operations. On a Kintex UltraScale+ FPGA, the CPI-optimized configuration achieves 1.221 cycles per instruction at 220 MHz, compared with 4.000 for an algorithmic HLS baseline and 1.575 at 242 MHz for an Ibex reference. An area-optimized configuration reduces LUTs by 65\% and registers by 47\%, with 9.7\% more benchmark cycles and a 250 MHz achieved clock. These results demonstrate a separation between designer-controlled architecture and tool-controlled scheduling.

cs.AR↗
arXiv2026

SpatialMem: Metric-Aligned Long-Horizon Video Memory for Language Grounding and QA

SpatialMem is a memory-based system which integrates 3D geometry, semantics, and language with one another in one combined, queryable representation. Starting with casually recorded egocentric RGB video, SpatialMem recreates metrically scaled interior scenes, recognizes structural 3D anchors (walls, doors, windows) as the first-layer scaffold, and associates open-vocabulary object nodes with evidence patches, visual embeddings, and two-layer descriptions to 3D coordinates, enabling spatial-relation reasoning, language-directed retrieval, and QA without special sensors. We test at scale on public benchmarks. On VSI-Bench ($288$ videos, $5{,}130$ questions) SpatialMem scores $37.5$ overall with RGB alone, and its margin is precisely where a metric memory should earn it (room size $41.4$ vs.\ $24.4$ with a same-size frame model). On the episodic-memory task of OpenEQA ($953$ questions) the memory serialised to text is lower than the blind control, but as evidence-frame retrieval it recovers frame-level accuracy, matching uniform sampling at an equal eight-frame budget and pulling ahead as the budget tightens, with every answer grounded in linked 3D evidence. On $12$ ScanNet scenes evaluated using instance-level official annotations, the anchor-first hierarchy assigns the retrieved object to the proper structural parent where open-vocabulary 3D scene graphs essentially cannot ($0.24$ compared to $0.01$ with matched input), a capability difference instead of an overall win. The queryable index has a cost of $14$\,KB per scene without the linked evidence frames, and we make available the suites, memories and code.

cs.CV↗
arXiv2026

RxnExBench: Benchmarking Vision-Language Models for Structured Chemical Reaction Extraction from Scientific Literature

Screen Tables and Substrate Scopes are information-dense figures widely used in synthetic chemistry literature, yet automated extraction of structured reaction data from them remains challenging. We introduce RxnExBench, a multimodal benchmark for fine-grained reaction-level evaluation. Built from high-impact chemistry journals, it contains 3,565 expert-annotated reactions and 2,361 molecular structures. The dataset adopts a unified schema with 15 reaction roles and 26 core fields, supporting downstream AI-for-Chemistry tasks such as retrosynthetic planning, yield prediction, and reaction condition recommendation. RxnExBench further provides a reasoning-driven difficulty taxonomy, a modular annotation pipeline, and the Oracle-MolRec protocol, which disentangles molecular recognition from higher-level reasoning. Evaluation of 9 state-of-the-art VLMs shows that performance degrades substantially as schema complexity and reasoning demands increase. Under the Oracle-MolRec setting, results further reveal that accurate molecular recognition alone is insufficient, with multimodal integration and chemistry-aware reasoning remaining the primary bottlenecks. We release all data and code to facilitate future research in chemistry-aware multimodal understanding and AI-for-Chemistry dataset construction. The dataset and code will be released at https://github.com/opendatalab/RxnExBench.

cs.AI↗
arXiv2026

In situ enhanced detectability of axions' responses in a microwave cavity

Microwave cavities in strong magnetic fields are among the most sensitive detectors for axion dark matter detection. Its sensitivity, however, is fundamentally limited by the second order nature of axion photon energy conversion, which yields an extremely weak signal. Here we propose a feasible approach to break this limitation, by pre-exciting the target cavity mode to generate a first order power response signal in situ. The resulting signal amplitude scales linearly with the axion-photon coupling parameter, which is in sharp contrast to the quadratic scaling of conventional schemes, and is detectable via standard quadrature balanced homodyne detection system, even though its phase remains stochastic. Numerical simulations demonstrate that this pre-excited seed field can effectively suppress thermal noise, and thus the detection sensitivity of the system can approach the lower quantum noise limit. Consequently, the sensitivities predicted by the KSVZ and DFSZ models could be surpassed. This proposal might offer a pathway to a major leap in axion dark matter searches, as it is within reach of current technology.

hep-ph↗
arXiv2026

OpenEarth-OVS44: A Multimodal Remote Sensing Dataset for Open-Vocabulary Segmentation

In remote sensing, open-vocabulary segmentation is particularly important for flexible and scalable land-cover understanding across diverse regions, as it improves generalization to novel categories without requiring exhaustive pixel-level annotations. Despite its great promise, progress in this area remains largely constrained by existing datasets with limited category coverage and overly coarse taxonomies, and therefore struggles to support fine-grained semantic understanding required in real-world applications such as urban functional zoning, carbon accounting, and post-disaster assessment. To address these challenges, we present OpenEarth-OVS44, a multimodal benchmark with 290,587 expert-verified segments from 2,832 RGB images and 1,341 strictly registered RGB-SAR pairs at sub-meter spatial resolution, covering 40 cities across France, Japan, and the United States under a unified 44-class taxonomy organized into 9 base categories. Three evaluation tracks are provided: an RGB-only fine-grained benchmark, a multimodal track with diverse synthetic cloudy conditions for weather-robust evaluation, and a held-out cross-region track. Extensive experiments on representative open-vocabulary segmentation models demonstrate that training on OpenEarth-OVS44 substantially improves fine-grained segmentation performance over the strongest training-free baseline, with improvements of up to 7.1 mIoU points on unseen categories and 27.6 mIoU points on seen categories. The dataset and code will be released.

cs.CV↗
arXiv2026

GeoNaviDiff: Transferring Disentangled Semantic Controls to Diffusion Features via Local Riemannian Navigation

Disentangled real-image manipulation requires factor-specific semantic control while preserving image fidelity and non-target content. StyleSpace $S$ provides interpretable controls but is limited by GAN inversion, whereas diffusion models offer stronger image realization without an explicit factorized control basis. We introduce GeoNaviDiff, an unsupervised cross-space framework that transfers disentangled StyleSpace interventions to local Riemannian trajectories in diffusion $h$-space. GeoNaviDiff constructs reconstruction-decoupled supervision by extracting a spherical Log-map tangent from a matched StyleGAN reconstruction-edit pair and parallel-transporting it to the real-image diffusion state. A shared, factor-conditioned Navigator predicts this tangent at inference, while norm-preserving geodesic realization and source-anchored rollout stabilize the reverse process without edited references or factor-specific models. Experiments across multiple datasets show that GeoNaviDiff outperforms the compared methods in fidelity, factor selectivity, edit locality, and trajectory smoothness.

cs.CV↗
arXiv2026

The Window Is in the Reasoning: When LLM Agents Decide to Follow Prompt Injections

LLM agents that read untrusted content can be hijacked by prompt injections. Runtime monitors that try to catch a hijack before the agent acts, and benchmarks that score each task with a single run, share an assumption: whether an agent follows an injection is settled when the injection is read. We test this assumption with branching rollouts that replay an episode exactly up to the step at which the agent reads an injection and then resample the rest of the episode many times, splitting the outcome variance into a part fixed at read time and a part decided afterwards. Across four AgentDojo suites, a Qwen3 agent without thinking is largely decided at read time (0.91 of the outcome variance lies between states); with thinking on, this share drops to 0.62 (95% CI of the drop [-0.38, -0.22]), in the same direction for three attack templates and two success labels. How much is fixed at read time is a property of the model that the same measurement reveals, and thinking lowers the attack success rate of Qwen3 but raises it for two other models. Resampling from reasoning prefixes shows that the decision forms gradually as the agent reasons. Zero-shot chain-of-thought monitors built on cheaper models miss much of what is already decided, whereas stronger monitors, and a lookahead monitor that judges simulated next actions, come close to what the prefix determines. Finally, single-run labels of thinking agents disagree with a rerun up to 35% of the time, so benchmarks need repeated runs per task.

cs.CL↗
arXiv2026

Fiber product of condensable algebras and critical points of boundary phase transitions

In this work, we propose a mathematical description of the critical point of boundary phase transitions in 2+1D topological orders. Given two boundary phases, $\mathcal{C}_A$ and $\mathcal{C}_B$, with corresponding Lagrangian algebras $A$ and $B$ in a 2+1D topological order $\mathcal{C}$, respectively, we propose that the critical point of the phase transition between these two phases, if it exists, corresponds to $D = A \times_{M} B$, which is the fiber product of $A$ and $B$ over $A$-$B$-algebra $M$. We prove that $D$ is also a condensable algebra. Additionally, we present an alternative classification of condensable algebras in $Z(Vec_G^ω)$ and reveal splitting phenomenon that arises during the condensation of certain algebras. We also perform explicit computations for the condensable algebras within several examples of $Z(Vec_G^ω)$, covering both Abelian and non-Abelian groups $G$.

math.CT↗
arXiv2026

Chemical-Potential-Dependent Superconducting Spectroscopy on Bi2Te3/NbSe2 Heterostructures

High-resolution scanning tunneling spectroscopy performed at dilution refrigerator temperatures was employed to investigate the superconducting properties of topological insulator/superconductor (Bi2Te3/NbSe2) heterostructures. Coherence peaks with inner kinks or inner peaks, deviating from the conventional single s-wave spectral line shape, were observed in both 5 quintuple-layer (5QL) and 7QL Bi2Te3 thin films. Based on the combination of chemical-potential-dependent spectroscopic measurements and model simulations, we conclude that these spectral features arise from contributions of bulk and surface states in Bi2Te3 by selectively proximity-coupled to a single superconducting gap from NbSe2, rather than inheriting the complicated superconductor's order parameter of NbSe2.

cond-mat.supr-con↗
arXiv2026

A Survey on LLM-Integrated Hardware Design Verification

Large language models (LLMs) are increasingly being integrated into hardware verification to automate specification interpretation, verification-artifact generation, debugging, formal reasoning, and tool orchestration. This survey provides a systematic review of LLM-assisted hardware functional verification across SystemVerilog assertion generation, stimulus and testbench generation, bug localization and design repair, model checking and equivalence checking, SAT/SMT optimization, and emerging agentic verification workflows. We organize the literature by methodology, verification objective, tool interaction, benchmark, and evaluation criterion, and examine both inference-time techniques--including prompting, retrieval, structured reasoning, and agentic workflows--and training-time adaptation. Across these areas, a common pattern emerges: LLMs are most effective as semantic reasoning, search, and orchestration components embedded within verification-aware workflows, while simulators, formal engines, coverage tools, and solvers provide executable feedback and correctness evidence. However, tool acceptance alone does not establish verification correctness, since assertions, tests, repairs, or proofs may satisfy available checks without faithfully capturing the complete design intent. We therefore identify semantic alignment between specifications and verification evidence, scalable integration with deterministic tools, generalization to unseen designs, and rigorous evaluation of correctness, cost, robustness, and human effort as key challenges. Finally, we discuss emerging directions toward specification-centered, neuro-symbolic, and persistent agentic verification systems that combine LLM flexibility with independently checkable verification evidence.

cs.AR↗
arXiv2026

REACT: Physically and Chemically Consistent Reconstruction of Marine Active Tracers

Reconstructing global sea surface pH from sparse observations is critical for monitoring ocean acidification and understanding marine carbon cycling. Traditional assimilation and inverse models are physically grounded but costly for large-scale reconstruction. Recent black-box and physics-guided AI models improve efficiency, but are mainly designed for passive tracers, where the reconstructed variable is also the transported inventory. In contrast, pH is an active carbonate tracer: it is the prediction target, while dissolved inorganic carbon (DIC) is the conserved carbon inventory. This mismatch can produce low pH error while violating carbonate closure and source-free carbon conservation. To address this, we introduce \textbf{REACT}, a carbon-first reconstruction framework that decouples transport, active correction, and chemical decoding. REACT transports a latent carbonate state with a conservative advection--diffusion solver, captures non-conservative carbon-cycle variations with a source module, decodes the corrected state into pH, and constrains the output through carbonate equilibrium. This design keeps pH as the target while enforcing consistency on the underlying carbon state. On simulation data, REACT reduces pH NRMSE by (14.7%) and chemical consistency error by (24.0%) over the best baseline. Cross-temporal-scale evaluations show robustness against error accumulation from coarse to fine temporal scales, and ablation studies validate the effectiveness of each component.

cs.AI↗
arXiv2026

Atomic-Scale Imaging of Lattice Relaxation and Topological Flat Bands in Helical Trilayer Graphene

Helical trilayer graphene (HTG) has emerged as a highly tunable moire quantum material that hosts strong electronic correlations and nontrivial band topology. However, the atomic-scale lattice structure and local electronic properties have remained largely unexplored. Here we present a comprehensive real-space study of HTG using a combination of scanning near-field optical microscopy and low-temperature scanning tunneling microscopy. We directly image supermoire lattice relaxation, revealing large triangular domains separated by sharp domain walls, as well as stripe domains connected by smoothly varying boundaries. Atomic-scale spectroscopy uncovers flat bands with a honeycomb electronic texture and one-dimensional boundary states confined to domain walls. By systematically varying the twist angle, we identify a magic angle of approximately 1.9°, substantially larger than the 1.6 degree predicted by theory. Our results establish a direct microscopic link between lattice relaxation and flat bands in HTG.

cond-mat.mes-hall↗
arXiv2026

AquaMend: Minimal Re-probing and Conditional Rollback for Latent-Belief Failures in Embodied Agents

Physical changes or sensing errors can invalidate embodied agents' task-relevant beliefs. AquaMend compares re-probing, rollback, and supported continuation on a probe-belief-action graph under an expected-loss objective covering sensing, physical recovery, and uncorrected failures. A joint posterior guides a one-step policy with conditional detection-power screening. The per-belief three-way optimum requires independence, separability, and fully resolving probes; the general policy has no global optimality guarantee. Across 32 paired scenarios in a self-constructed simulation benchmark, AquaMend recovers in 28/32 cases and reduces mean complete loss by 21.6% versus restart. Its paired loss difference from decision-theoretic troubleshooting (DTT) is not statistically significant after Holm correction. Against the all-candidate ablation, online decision time decreases by 12.3% overall but increases by 3.4% in the uncovered late stage.

cs.RO↗
arXiv2026

Safety Does Not Compose: Non-Decaying Loop State for Autonomous LLM Agents

Large language model agents are increasingly deployed as autonomous loops. Starting from one human goal, such a system repeatedly discovers work, plans, executes tool calls, verifies outcomes and persists state across many unattended iterations. The agent safeguards in wide use, however, are defined over a single trajectory, and their safety state is re-initialized when the next trajectory begins. We show that this is a failure of composition rather than an implementation detail. Our central result is a separation: against an attack whose evidence is fragmented across several iterations, every trajectory-scoped monitor has a true-positive rate equal to its false-positive rate, however expressive it is, because the evidence it would need never appears in the window it sees, whereas a monitor retaining cross-iteration state separates the two perfectly. We further show that the obvious repair of carrying a geometrically decaying risk score is insufficient, because the cooling-off period a patient adversary must wait is a constant that does not grow with the horizon $N$. We then present LoopHarness, which restores a persistent, non-decaying safety state at the loop level. Under mediated commits and an arbiter detection floor $δ_M$, it bounds the expected number of unauthorized irreversible actions by $B+m-1+m/δ_M$, a constant in $N$, of which the $B+m-1$ term is decided by a model-free rule and therefore survives a fully colluding verifier. We give a complete evaluation protocol on native Agent-SafetyBench tasks with paired clean and attacked episodes, an outer-state attack suite whose decisive evidence exists only across iterations, per-module ablations, and an adaptive white-box red team.

cs.CR↗
arXiv2026

To Memories and Beyond: From Remembering to Knowing You across Long-Term Multimodal Personal Archives

As AI systems evolve into personalized digital companions, a central capability is reasoning over a user's long-term personal history: not merely storing past events, but tracking longitudinal experiences and evolving preferences. Progress here is bottlenecked by evaluation, existing long-term memory benchmarks are largely synthetic and text-only, they overlook the visual records that anchor everyday human memory, lack the authentic and causally connected longitudinal data that real personalization demands, and consequently remain confined to shallow factual recall. We introduce ReaLMem (Real-world Long-term Multimodal Memory), the first benchmark built from authentic multi-year personal visual archives, paired with first-person subjective annotations. ReaLMem evaluates models across three cognitive tiers of increasing difficulty: factual recall, persona inference, and predictive personalization. We further propose ChronoProfiler, a temporal-weighting profiling module that computes temporal stability scores for user attributes and applies them as a salience prior, resolving conflicts among temporally inconsistent preferences and helping models compound multiple co-active preferences in complex personalized decisions. Extensive evaluation of frontier multimodal large language models (MLLMs) and memory systems on ReaLMem reveals predictive personalization as a consistent ceiling, exposes clear performance gaps and bottlenecks between MLLMs and memory systems, and shows that high-quality, temporally informed representations substantially improve personalization. Together, ReaLMem and ChronoProfiler provide an authentic testbed and a simple, effective mechanism for long-term personalization, laying a foundation for future research on lifelong AI companions.

cs.CL↗
arXiv2026

Harness-agnostic detection and immunization of reward hacking in self-evolving language models

Self-evolving language models improve by proposing candidate updates and keeping whatever raises a visible score. When that score is an imperfect proxy for the capability one actually wants, sustained selection widens the gap between the two. This is reward hacking. We introduce HackProbe, a monitor that attaches to an arbitrary self-evolving loop through two black-box hooks, with no access to weights or activations. It keeps a secret, distribution-fixed comparison core, whose frozen distribution makes its capability proxy comparable across generations, alongside a rotated fresh layer that hardens the bank against co-adaptation. Four tests built on that proxy cover the level gap, a scale-aligned divergence with online change-point detection, capability stagnation, and a conditional confidently-wrong rate; a Sidak correction turns them into a calibrated family-wise p-value. Diagnosis alone recovers nothing, so a risk-aware immunization layer reselects an honest candidate from the proposal pool using the core together with a purely structural gaming footprint, disclosing at most log2 Pi bits per generation to the host. We prove a detectability bound that converts a target error rate into an explicit probe-size budget, and we delimit what probe rotation does and does not buy. On a controlled prompt-level host with four injected hacking channels and ground-truth labels, HackProbe reaches 0.763 AUROC against 0.663 for the strongest baseline and cuts the false-positive rate from 0.706 to 0.434. Its bandwidth-limited reselection is the only immunization level that returns more true capability under hacking, 5.2 points on average, than it forfeits on clean runs, 4.7; per-channel effects are mostly not individually significant.

cs.AI↗
arXiv2026

Site-specific Channel Modeling Based on Remote-Sensing Maps for 6G Space--Air--Ground Digital Twins

Site-specific channel models are essential for wireless digital twins of 6G space--air--ground communication systems. However, 3D maps are difficult to obtain over wide areas, which limits large-area site-specific channel modeling. To address this issue, this paper proposes a remote-sensing-based augmented ray-tracing channel modeling framework. The framework comprises a deterministic RT branch, a measurement-statistical branch, and an RT augmentation branch. To overcome the difficulty of acquiring large-area 3D maps, the deterministic RT branch reconstructs a 3D RT scene from satellite remote-sensing imagery and calibrates its electromagnetic material parameters using measured path loss. To provide the statistical parameters required for RT augmentation, the measurement-statistical branch establishes the marginal distributions and interparameter dependence models of the channel parameters. Specifically, a wideband UAV channel measurement campaign is conducted at 4.60 GHz, and a proposed multipath estimation method estimates the complex amplitudes, delays, and Doppler shifts of the measured multipath. To bridge the gap between RT predictions and measurements, the RT augmentation branch organizes the RT multipath into LoS, LoS-tail, and NLoS components, generates additional short-delay LoS-tail paths, and reallocates the component and path powers according to the measurement-derived statistics while preserving the total RT received power. The validation results show that the proposed framework reduces the path loss RMSE from 5.45 to 4.35 dB and, relative to calibrated RT, decreases the RMS delay spread and normalized Doppler spread RMSEs by 53.03 and 26.48, respectively. The proposed framework provides a site-specific channel modeling approach for 6G space--air--ground digital-twin studies.

eess.SP↗