arXiv subjects

Haoran Zhao

Publications and source records attributed to Haoran Zhao.

At least 19 recordsLinked to original sources

arXiv2026

CLOSER-VLN: Closed-Loop Self-Verified Retrieval-Augmented Reasoning for Aerial Vision-Language Navigation

Vision-language navigation (VLN) has recently advanced with large language and multimodal models, enabling agents to follow natural-language instructions in unseen environments without training a task-specific navigation policy. However, most existing VLN methods relying on large models still adopt an open-loop decision-execution approach, where candidate actions are generated from instructions and observations but are rarely verified or corrected before execution. This causes critical issues in aerial VLN, where minor errors in intermediate actions may quickly accumulate into large trajectory deviations and lead to target loss. To address this issue, we propose Closed-loop Self-verified Retrieval-augmented Reasoning (CLOSER), a training-policy-free method that sequentially performs action reasoning, reliability verification, targeted retrieval, and action correction in a closed-loop manner before executing concrete actions. We instantiate the CLOSER in aerial VLN tasks and develop a CLOSER-VLN framework, which is composed of three components: a hierarchical reasoner for generating candidate actions based on available information, a multidimensional action verifier for assessing the reliability of actions generated by the reasoner, and a verification-triggered multimodal retriever for retrieving targeted exemplars from a memory bank only when verification fails. We conduct experimental evaluations on the CityNav benchmark, where CLOSER-VLN achieves 32.01% SR and 21.28% SPL on the test-unseen split, confirming the effectiveness of closed-loop reasoning.

cs.CV↗

arXiv2026

When Correct Decisions Hide Internal Stress: Decision-State Probing in Multimodal Language Models

Multimodal language models are typically evaluated through external behavior: selecting the correct image--text match, rejecting unsupported captions, or answering visual queries correctly. However, correct behavior alone does not show that the model's internal decision state remains stable under controlled semantic stress. We study this gap through S$^3$E (Structured Semantic Stress Evaluation), a framework for analyzing behavior-internal decoupling in multimodal language models. S$^3$E uses a positive-anchored A/B forced-choice setup in which an image-supported caption is contrasted against semantic stress candidates under both original and swapped option orders, while hidden states are extracted at the pre-answer decision state. We focus on strict-correct trials, where the model consistently selects the correct caption across both orders. Rather than treating arbitrary hidden-state variation as evidence of instability, we measure whether semantic-conflict candidates induce excess decision-state displacement relative to meaning-preserving controls. Across Qwen3VL, Gemma3, and InternVL3, semantic stress consistently produces positive selected-layer excess displacement over lexical controls despite correct forced-choice behavior, while comparisons against random negatives are model-dependent. We interpret this as a scoped decision-state stress-sensitivity signal rather than evidence of downstream failure or hallucination. Our results suggest that forced-choice correctness alone is not a sufficient certificate of invariant internal decision geometry.

cs.CL↗

arXiv2026

Symmetry Breaking as Quantum Gate: Entropy and Weak Mixing Angle

We establish a correspondence between two independent entropic probes -- the variation of R\'{e}nyi mutual information (RMI) across the electroweak symmetry breaking (EWSB) transition and the stabilizer R\'enyi entropy (SRE) -- in tree-level $2\to 2$ elastic scatterings. After angular averaging, the RMI (helicity basis) and the SRE (fixed beam basis) exhibit identical dependence on $\sin^2\theta_W$ within each neutral-current channel. We trace this correspondence to a common physical origin that it's the Yukawa mass insertion acts as a $-\mathrm{i}Y$ quantum gate in chirality space. Minimizing entropies across all processes yields $\sin^2\theta_W$ values matching purely axial vector-like couplings in $Z$ boson exchanged channel.

hep-ph↗

arXiv2026

LightZeroNav: Zero-Shot Vision Language Navigation in Continuous Environments Based on Lightweight VLMs

Although vision-language navigation (VLN) has progressed rapidly, zero-shot VLN in continuous environments (VLN-CE) remains highly challenging when using lightweight vision-language models (VLMs), whose limited reasoning capacity makes long-horizon navigation unreliable. In this paper, we propose LightZeroNav to tackle the three major bottlenecks when using lightweight VLMs in zero-shot VLN-CE,i.e.,information redundancy from multi-source inputs, inaccurate progress estimation caused by noisy textual memory, and task entanglement between action execution and stage transition. Using only RGB observations and a lightweight open-source Qwen3-VL-8B backbone, LightZeroNav achieves competitive performance with GPT-4o (~200B) without task-specific training, graph search, or waypoint predictors, demonstrating its effectiveness in zero-shot VLN-CE.

cs.CV↗

arXiv2026

PM-Nav: Priori-Map Guided Embodied Navigation in Functional Buildings

Existing language-driven embodied navigation paradigms face challenges in functional buildings (FBs) with highly similar features, as they lack the ability to effectively utilize priori spatial knowledge. To tackle this issue, we propose a Priori-Map Guided Embodied Navigation (PM-Nav), wherein environmental maps are transformed into navigation-friendly semantic priori-maps, a hierarchical chain-of-thought prompt template with an annotation priori-map is designed to enable precise path planning, and a multi-model collaborative action output mechanism is built to accomplish positioning decisions and execution control for navigation planning. Comprehensive tests using a home-made FB dataset show that the PM-Nav obtains average improvements of 511\% and 1175\%, and 650\% and 400\% over the SG-Nav and the InstructNav in simulation and real-world, respectively. These tremendous boosts elucidate the great potential of using the PM-Nav as a backbone navigation framework for FBs.

cs.RO↗

arXiv2026

ViSA-Enhanced Aerial VLN: A Visual-Spatial Reasoning Enhanced Framework for Aerial Vision-Language Navigation

Existing aerial Vision-Language Navigation (VLN) methods predominantly adopt a detection-and-planning pipeline, which converts open-vocabulary detections into discrete textual scene graphs. These approaches are plagued by inadequate spatial reasoning capabilities and inherent linguistic ambiguities. To address these bottlenecks, we propose a Visual-Spatial Reasoning (ViSA) enhanced framework for aerial VLN. Specifically, a triple-phase collaborative architecture is designed to leverage structured visual prompting, enabling Vision-Language Models (VLMs) to perform direct reasoning on image planes without the need for additional training or complex intermediate representations. Comprehensive evaluations on the CityNav benchmark demonstrate that the ViSA-enhanced VLN achieves a 70.3\% improvement in success rate compared to the fully trained state-of-the-art (SOTA) method, elucidating its great potential as a backbone for aerial VLN systems.

cs.CV↗

arXiv2026

On Theoretical Stability Proof and Stability Margin Analysis of Enhanced Droop-Free Control Schemes for Islanded Microgrids

This paper studies enhanced droop-free control strategies with sparse neighboring communication for achieving effective active power sharing of distributed energy resources (DERs) while maintaining the frequency stability of islanded microgrids. The normalized active power consensus (NAPC) based droop-free control can share the load among controllable DERs in proportion to their available capacities. However, existing literature exclusively takes the asymptotic stability of the NAPC based droop-free control for granted, lacking a comprehensive theoretical proof that is critical for ensuring its effective design and practical implementation. This paper, for the first time, provides a thorough theoretical proof of the asymptotic stability of two NAPC-based droop-free control schemes: ordinary NAPC (ONAPC) and amplifier-equipped NAPC (A-NAPC), by testifying that all effective eigenvalues have negative real parts. The effect of various system settings on the stability margins is further analyzed with respect to the average admittance of the electrical network, the sparseness of the communication network, and the average available capacity of controllable DERs. Based on the sensitivity of eigenvalues with respect to perturbations, a vulnerability analysis is conducted to identify the weaknesses in the microgrids. Case studies demonstrate that the available capacity of controllable DERs has the most decisive influence on the stability margin of NAPC-based droop-free control, while O-NAPC/ANAPC control scheme is more suitable for microgrids with DERs of larger/ smaller available capacities.

eess.SY↗

arXiv2026

CFHT MegaCam Two Deep Fields Imaging Survey (2DFIS) II: Decoding the Lensing Profile of a "Rotating" Cluster with Deep CFHT Imaging

We present a multi-wavelength analysis of the galaxy cluster RXCJ0110.0+1358 ($z=0.058$), a rotating cluster candidate, combining deep CFHT imaging, SDSS photometry, spectroscopic redshifts, and XMM-Newton X-ray observations. We find a notable discrepancy between the optical and X-ray views: while optical data reveal a pronounced bimodal galaxy distribution with significant kinematic substructure signatures, the X-ray emission exhibits a single, smoothly extended component centered on the BCG. Our weak lensing analysis resolves this discrepancy by revealing that the mass is predominantly concentrated in the southeast ($\log M_{200}/M_\odot = 14.04_{-0.40}^{+0.24}$), while the northwestern substructure has a negligible mass ($\sim 10^{13} M_\odot$). This immense mass disparity rules out the dynamical possibility of a rotating system. We demonstrate that the apparent optical bimodality arises from the projection of a filament, which led optical group-finding algorithms to misclassify these galaxies as cluster members. This contamination creates a spurious substructure that mimics a rotation signal and leads to an overestimation of the luminosity-based halo mass, resolving the observed inconsistencies.

astro-ph.GA↗

arXiv2026

When Is Rank-1 Enough? Geometry-Guided Initialization for Parameter-Efficient Fine-Tuning

Parameter-efficient fine-tuning (PEFT) is a standard way to adapt multimodal large language models, yet extremely low-rank settings -- especially rank-1 LoRA -- are often unstable. We show that this instability is not solely due to limited capacity: in the rank-1 regime, optimization is highly sensitive to the update direction. Concretely, pretrained vision and text features form mismatched anisotropic regions, yielding a dominant "gap" direction that acts like a translation component and disproportionately steers early gradients under rank-1 constraints. Analyzing pretrained representations, we identify a modality-gap axis that dominates early gradient flow, while a random rank-1 initialization is unlikely to align with it, leading to weak gradients and training collapse. We propose Gap-Init, a geometry-aware initialization that aligns the rank-1 LoRA direction with an estimated modality-gap vector from a small calibration set, while keeping the initial LoRA update zero. Across multiple vision-language tasks and backbones, Gap-Init consistently stabilizes rank-1 training and can match or outperform strong rank-8 baselines. Our results suggest that at the extreme low-rank limit, initial alignment can matter as much as rank itself.

cs.LG↗

arXiv2025

Ultrafast electron diffractive imaging of the dissociation of pre-excited molecules

Gas phase ultrafast electron diffraction (GUED) has become a powerful technique to directly observe the structural dynamics of photoexcited molecules. GUED reveals information about the nuclear motions that is complementary to the information on the electronic states provided by spectroscopic measurements. GUED experiments so far have utilized a single laser pulse to excite the molecules and an electron pulse to probe the dynamics. This limits the excited states which can be studied to only those that can be reached by absorption of a photon from the ground state or in some cases simultaneous absorption of multiple photons. A broader class of experiments and dynamics can be accessed using two time-delayed laser pulses to access unexplored regions of the potential energy surfaces. As a proof-of-principle experiment using a double excitation, we studied the photodissociation of trifluoroiodomethane molecules that are pre-excited with an infrared (800 nm) femtosecond laser pulse before photo-dissociation is triggered with an ultraviolet (266 nm) femtosecond laser pulse. We have observed significant differences in the dissociation dynamics, with pre-excitation resulting in a slower dissociation process. This new capability can offer new insights on the evolution of nuclear wavepackets in regions of the excited potential energy surface which are inaccessible in single photon excitation. We present a methodology to carry out the measurement, analyze and interpret the data that could be applied to a broad class of experiments.

physics.chem-ph↗

arXiv2025

Analytical modeling of a stop-less modular bus line: Optimization, feasibility, and economies of scale

Conventional bus services often struggle with inefficiencies including prolonged dwell times at heavily used stops, especially for through passengers. A stop-less autonomous modular bus service (SLAM) has been proposed to reduce dwell times by decoupling the front pod to serve stops and then coupling it to the next bus. However, the optimal service design and feasibility region remain underexplored, despite their importance for planning and deployment. We propose an analytical optimization model that characterizes the optimal design, feasibility conditions, and sources of scale economies. Three novel constraints distinguish SLAM from conventional bus services: (i) a minimum headway to ensure sufficient time for decoupling, alighting, boarding, and coupling operations, (ii) a maximum headway to guarantee all passengers arriving within a headway fit in the standby pod, and (iii) a minimum bus length constraint, requiring at least two pods per bus to run in a SLAM manner. As ridership grows, the optimal design evolves through several regimes, in which headway constraints alternate between slack and binding states, while capacity constraints shift from one active form to another. Our analysis indicates that, compared with conventional services, SLAM is most suitable at intermediate demand levels: at low demand, the fixed costs of standby pods and the minimum two-pod configuration outweigh the time-saving benefits, whereas at high demand, non-stopping operation becomes infeasible. We further decompose the sources of scale economies into four components: the Mohring effect, through-capacity economies, boarding-capacity economies, and standby-pod costs, identifying under which conditions each of them is present. The numerical results validate the theoretical analysis.

eess.SY↗

arXiv2025

PhoenixCodec: Taming Neural Speech Coding for Extreme Low-Resource Scenarios

This paper presents PhoenixCodec, a comprehensive neural speech coding and decoding framework designed for extremely low-resource conditions. The proposed system integrates an optimized asymmetric frequency-time architecture, a Cyclical Calibration and Refinement (CCR) training strategy, and a noise-invariant fine-tuning procedure. Under stringent constraints - computation below 700 MFLOPs, latency less than 30 ms, and dual-rate support at 1 kbps and 6 kbps - existing methods face a trade-off between efficiency and quality. PhoenixCodec addresses these challenges by alleviating the resource scattering of conventional decoders, employing CCR to enhance optimization stability, and enhancing robustness through noisy-sample fine-tuning. In the LRAC 2025 Challenge Track 1, the proposed system ranked third overall and demonstrated the best performance at 1 kbps in both real-world noise and reverberation and intelligibility in clean tests, confirming its effectiveness.

eess.AS↗

arXiv2025

Typhoon Tracks Regulated by Feedbacks of Fine-Scale Clouds to Environment

Accurate tropical cyclone (TC) track prediction is crucial for mitigating the catastrophic impacts of TCs on human life and the environment. Despite decades of research on tropical cyclone (TC) track prediction, large errors known as track forecast busts (TFBs) occur frequently, and their causes remain poorly understood. Here, we examine a few dozens of TCs using a unique TC downscaling strategy that can quantitatively assess the sensitivity of TC track on the strength of feedbacks of fine-scale clouds to environment. We show that as TFBs have a weaker environmental steering that favors scattering cumulonimbus clouds, capturing asymmetric distribution of planetary vorticity advection induced by such fine-scale clouds corrects TFBs by 60 percent. Our clear identification of such important TC track predictability source promises continuous improvement of TC track prediction as finer-scale TC clouds and their interactions with environment are better resolved as model larger-scale behaviors have improved.

physics.ao-ph↗

arXiv2025

GSM8K-V: Can Vision Language Models Solve Grade School Math Word Problems in Visual Contexts

Mathematical reasoning is a key capability for vision-language models (VLMs), yet current benchmarks mainly evaluate text-based or explicitly symbolic visual inputs. It remains unclear whether VLMs can reason mathematically when information must be perceived and inferred from images rather than read from explicit symbols. We introduce GSM8K-V, a benchmark transforming GSM8K into multi-image sequences with semantic equivalence preserved. By mapping text-based problems into visual form via an automated pipeline and human verification, we curate 1,319 high-quality samples. In GSM8K-V, quantities must be extracted through visual perception, and reasoning chains must be reconstructed by integrating implicit cues across scenes. Evaluation of 34 VLMs reveals a striking modality gap: while most models exceed 90\% on text, the best model achieves only 59\% on GSM8K-V, far below the 91\% human accuracy. Notably, models enhanced for visual math reasoning show no improvement on GSM8K-V despite large gains on existing benchmarks, confirming that it evaluates a distinct capability. Error analysis shows that the primary bottleneck lies in Implicit Visual Inference Error (IVIE), where models fail to recover visual semantics that are implied rather than explicitly stated. Our code and data are released at https://github.com/ZJU-REAL/GSM8K-V.

cs.CV↗

arXiv2025

SAGE: Semantic-Aware Shared Sampling for Efficient Diffusion

Diffusion models manifest evident benefits across diverse domains, yet their high sampling cost, requiring dozens of sequential model evaluations, remains a major limitation. Prior efforts mainly accelerate sampling via optimized solvers or distillation, which treat each query independently. In contrast, we reduce total number of steps by sharing early-stage sampling across semantically similar queries. To enable such efficiency gains without sacrificing quality, we propose SAGE, a semantic-aware shared sampling framework that integrates a shared sampling scheme for efficiency and a tailored training strategy for quality preservation. Extensive experiments show that SAGE reduces sampling cost by 25.5%, while improving generation quality with 5.0% lower FID, 5.4% higher CLIP, and 160% higher diversity over baselines.

cs.LG↗

arXiv2025

A.S.E: A Repository-Level Benchmark for Evaluating Security in AI-Generated Code

The increasing adoption of large language models (LLMs) in software engineering necessitates rigorous security evaluation of their generated code. However, existing benchmarks often lack relevance to real-world AI-assisted programming scenarios, making them inadequate for assessing the practical security risks associated with AI-generated code in production environments. To address this gap, we introduce A.S.E (AI Code Generation Security Evaluation), a repository-level evaluation benchmark designed to closely mirror real-world AI programming tasks, offering a comprehensive and reliable framework for assessing the security of AI-generated code. Our evaluation of leading LLMs on A.S.E reveals several key findings. In particular, current LLMs still struggle with secure coding. The complexity in repository-level scenarios presents challenges for LLMs that typically perform well on snippet-level tasks. Moreover, a larger reasoning budget does not necessarily lead to better code generation. These observations offer valuable insights into the current state of AI code generation and help developers identify the most suitable models for practical tasks. They also lay the groundwork for refining LLMs to generate secure and efficient code in real-world applications.

cs.SE↗

arXiv2025

SE-VLN: A Self-Evolving Vision-Language Navigation Framework Based on Multimodal Large Language Models

Recent advances in vision-language navigation (VLN) were mainly attributed to emerging large language models (LLMs). These methods exhibited excellent generalization capabilities in instruction understanding and task reasoning. However, they were constrained by the fixed knowledge bases and reasoning abilities of LLMs, preventing fully incorporating experiential knowledge and thus resulting in a lack of efficient evolutionary capacity. To address this, we drew inspiration from the evolution capabilities of natural agents, and proposed a self-evolving VLN framework (SE-VLN) to endow VLN agents with the ability to continuously evolve during testing. To the best of our knowledge, it was the first time that an multimodal LLM-powered self-evolving VLN framework was proposed. Specifically, SE-VLN comprised three core modules, i.e., a hierarchical memory module to transfer successful and failure cases into reusable knowledge, a retrieval-augmented thought-based reasoning module to retrieve experience and enable multi-step decision-making, and a reflection module to realize continual evolution. Comprehensive tests illustrated that the SE-VLN achieved navigation success rates of 57% and 35.2% in unseen environments, representing absolute performance improvements of 23.9% and 15.0% over current state-of-the-art methods on R2R and REVERSE datasets, respectively. Moreover, the SE-VLN showed performance improvement with increasing experience repository, elucidating its great potential as a self-evolving agent framework for VLN.

cs.CV↗

arXiv2025

Comparing human and LLM politeness strategies in free production

Polite speech poses a fundamental alignment challenge for large language models (LLMs). Humans deploy a rich repertoire of linguistic strategies to balance informational and social goals -- from positive approaches that build rapport (compliments, expressions of interest) to negative strategies that minimize imposition (hedging, indirectness). We investigate whether LLMs employ a similarly context-sensitive repertoire by comparing human and LLM responses in both constrained and open-ended production tasks. We find that larger models ($\ge$70B parameters) successfully replicate key preferences from the computational pragmatics literature, and human evaluators surprisingly prefer LLM-generated responses in open-ended contexts. However, further linguistic analyses reveal that models disproportionately rely on negative politeness strategies even in positive contexts, potentially leading to misinterpretations. While modern LLMs demonstrate an impressive handle on politeness strategies, these subtle differences raise important questions about pragmatic alignment in AI systems.

cs.CL↗