arXiv subjects

Kun Wang

Publications and source records attributed to Kun Wang.

At least 19 recordsLinked to original sources

arXiv2026

Atmospheric Turbulence Profiling in the Lenghu Site Area Using MERRA-2 Data

The Lenghu site, in the Saishiteng Mountains of the northern Tibetan Plateau, has been identified as an excellent astronomical site from preliminary statistics of clear nights, sky background brightness, precipitable water vapour, total seeing and weather. Since 2022, several projects with apertures larger than 4 m have started construction or are being planned. Given the use of adaptive optics (AO) on these projects, in-depth measurements and analysis of the atmospheric turbulence above the site are required: the distribution and variation of the turbulence profiles provide a large-scale, climatological basis for site planning and for comparing candidate locations, complementing rather than replacing the on-site measurements needed for instrument design and operations. Predicting the atmospheric profiles of the site and its surroundings from satellite data, combined with on-site measurements, can help to optimise the site planning. This paper uses MERRA-2 reanalysis data to study the atmospheric profiles of the site and its environment, revealing the seasonal and long-term variations of the observing conditions. The reanalysis results are cross-checked against on-site ground-level instruments serving the same purpose. MERRA-2 overestimates the median free-atmosphere seeing by +0.22 arcsec (+50 %) while keeping the night-to-night correlation weak (r ~ 0.2), and underestimates the median total seeing by -0.24 arcsec (-27 %). It captures only ~22 % of the ground-layer turbulence and little of the observed seasonal amplitude of the free atmosphere. MERRA-2 is therefore suited to relative comparisons of spatial variability between locations and to climatological characterisation, not to quantitative AO applications or night-by-night assessment of profiles.

astro-ph.IM↗

arXiv2026

Interface Control of Second-Harmonic Generation in Layered CuInP2S6 via PbZr0.2Ti0.8O3 Domain Wall

The control of nonlinear optical responses in van der Waals ferroelectrics at the nanoscale offers opportunities for developing two-dimensional integrated nanophotonics with programmable functionalities. In this work, we report the first realization of nanoscale nonvolatile control of the amplitude and polarization of second-harmonic generation (SHG) response in layered CuInP2S6 (CIPS) via the domain structure in an interfacial PbZr0.2Ti0.8O3 (PZT) thin film. The SHG intensity at CIPS/PZT heterointerface is continuously modulated, either substantially enhanced or almost entirely quenched, by the circular domain walls in underlying ferroelectric oxide. The interfacial coupling of CIPS polar axis with in-plane polarization of PZT domain walls tailors the SHG light polarization into two-fold (C2V) symmetry, which is well modeled by nonlinear electromagnetic theory. Overall, this study enables nonvolatile programmable design of nonlinear light intensity and polarization in van der Waals ferroelectrics at the nanoscale.

cond-mat.mtrl-sci↗

arXiv2026

Scaling Qualitative Analysis with Large Language Models: Development and Validation of the Smart Coder Framework for Human-LLM Collaborative Coding

Current uses of large language models (LLMs) for qualitative coding often rely on ad hoc prompting with limited guidance for reliability and human oversight. To address these limitations, we propose the Smart Coder Framework, a three-phase approach structuring human-LLM collaborative coding through Instruction, Prompt Strategy, and Reliability Check. We evaluated the framework using 72 middle-school mathematics teacher interview transcripts and a human-consensus benchmark. A factorial design compared instructed versus uninstructed prompts, few-shot versus few-shot plus chain-of-thought prompting, and single-pass versus self-consistency coding. Twelve transcripts were used for prompt development and calibration, and 60 for the formal evaluation. In the results, instruction consistently improved agreement. Chain-of-thought further improved agreement under adequate instruction, and self-consistency produced smaller, theme-dependent gains but helped identify unstable decisions for human review. Performance also varied by transcript characteristics. Overall, the framework systematically structures and monitors LLM-assisted qualitative coding while preserving human authority over coding decisions.

cs.AI↗

arXiv2026

IQuest-Q1: Advancing Foundation Capabilities for Agentic CLI Systems

We present IQuest-Q1, a sparse Mixture-of-Experts (MoE) agentic foundation model with 320B total parameters (15B activated). IQuest-Q1 is developed through a three-stage process: (1) stable and efficient pre-training, combining Stable SplitNorm, Muon, and muP with a two-stage curriculum whose second stage increases the coding-data mixture, preserving a wide stability margin between safe and optimal learning rates; (2) fine-grained mid-training, expanding the data mixture across domains and formats and progressively extending the context; and (3) multi-scaffold with multi-version post-training, combining SFT, domain-specialized RL, and MOPD across various scaffolds to integrate complementary training signals and strengthen agentic coding capabilities. Furthermore, model merging combines different experts and checkpoints from different stages while mitigating conflicts among their parameter updates. The resulting IQuest-Q1 demonstrates capabilities in both coding and general tasks. We release IQuest-Q1 as an open-source model to support community research and development.

cs.CL↗

arXiv2026

In-Context Learning Can Help Vision-Language Models Overcome Training Priors

Vision-language models (VLMs) learn strong statistical regularities during training, which can make them fail to properly respond to visual evidence when input images violate those regularities. It is challenging to disambiguate these prior-induced errors from errors caused by the model missing the underlying visual capabilities. In this work, we use controlled visual in-context learning to encourage the model to focus on the visual evidence, and demonstrate that this alone is frequently sufficient to correct the errors -- suggesting that the models in fact do possess the relevant visual capabilities. Specifically, we show that across five vision-centric tasks, demonstrations yield large gains on prior-conflicting counterfactual examples ($+27.2\%$ for Qwen3-VL 235B and $+15.1\%$ for Qwen3-VL 32B) while leaving accuracy on prior-aligned examples nearly unchanged. These findings generalize across five VLM families with gains from $+9.7\%$ to $+30.3\%$. We analyze visual representations and find that demonstrations indeed redirect attention toward grounding cues while reducing reliance on cues that support the learned prior. Finally, we find that in-context learning is limited when the underlying visual capability is weak or absent. Together, these results show that current VLMs possess visual grounding capabilities that standard evaluations may overlook, underscoring the need for controlled evaluation design in diagnosing and improving model capability.

cs.CV↗

arXiv2026

DuplexLedger: Benchmarking Transactional Memory in Full-Duplex Speech Agents

Full-duplex speech agents must update persistent memory while listening and speaking, yet it remains unclear which events should be stored, revised, or excluded. Existing benchmarks isolate interaction timing from memory outcomes, obscuring the state transitions behind a correct answer. We introduce DuplexLedger, a bilingual benchmark of 1,094 persistent sessions aligning four full-duplex scenarios with event-level Write/Update/Ignore/Retain supervision, explicit pre/post memory states, and 6,564 same-session questions and checks. Evaluating five full-duplex models reveals an interaction-memory gap: no system combines strong boundary control with coherent final-state memory, and the best final-state fact coverage is 2.01%. We propose ACT-Mem, an 8.92M-parameter sidecar comprising an event controller, a versioned transactional ledger, and a typed retriever coupled with a frozen MiniCPM-o 4.5 backbone. On 219 held-out sessions, ACT-Mem increases latest-value accuracy from 4.11% to 8.68% while reducing stale assertions and distractor adoption. It improves interaction boundaries but slows interruption stopping. These results motivate selective, auditable full-duplex memory rather than indiscriminate transcript retention. Code is available at https://github.com/lyx-researcher/DuplexLedger.

cs.AI↗

arXiv2026

Metric Depth Estimation from Arbitrarily Degraded Low-Resolution Depth Prompts

We propose AdaDS, an Adaptive Diffusion Sampling framework for prompted metric depth estimation, which estimates high-resolution metric depth from an image and an arbitrarily degraded low-resolution depth prompt. This setting is commonly studied as depth super-resolution, where existing methods typically regress depth directly without an explicit prior over high-quality depth, and thus often produce artifacts under severe or unseen degradations. In contrast, AdaDS exploits the contraction property of Gaussian smoothing: as noise accumulates in the forward diffusion process, the noisy distributions of degraded prompts and their high-quality counterparts converge toward an isotropic Gaussian prior, while the prompt content is gradually erased. Accordingly, AdaDS refines the prompt and estimates its uncertainty to adaptively select a starting timestep, and injects learned noise to place the intermediate sample in a high-density region of the depth distribution learned by a pre-trained diffusion model. Denoising from this sample yields high-quality depth with fine-grained details and corrects residual errors, even under imperfect prompt refinement. Experiments on real-world and synthetic benchmarks demonstrate its superior zero-shot generalization and robustness to diverse degradations.

cs.CV↗

arXiv2026

Aerial-R1: Reinforcing Aerial Reasoning and Segmentation in MLLMs

While Multimodal Large Language Models (MLLMs) have revolutionized general visual perception, existing methods suffer from severe hallucination in the aerial domain, where extreme scale variations and visual ambiguity abound. Existing Supervised Fine-Tuning (SFT) paradigms force models to mimic ground truth without modeling uncertainty, often resulting in blind guessing on tiny or unrecognizable objects in aerial scenes. In this paper, we present Aerial-R1, the first aerial MLLM framework to reinforce visual reasoning capabilities for aerial reasoning and segmentation. Drawing inspiration from the reasoning-incentivizing mechanisms of DeepSeek-R1, we find that trustworthy perception stems from consistency across views, establishing a visual parallel to the Chain-of-Thought in logic. To effectively adapt policy optimization for non-autoregressive mask generation, we introduce Hybrid-View Group Advantage-Weighted Regression (H-GAWR). This strategy constructs dynamic groups of high-resolution and low-resolution views, seamlessly bridging group-relative advantage estimation with advantage-weighted policy extraction. By enforcing consensus on clear inputs and strictly penalizing hallucinations on ambiguous ones, Aerial-R1 develops an explicitly incentivized refusal mechanism. Extensive experiments on VSAI-Ref, the first aerial MLLM benchmark dedicated to jointly evaluating dense reasoning and hallucination robustness, demonstrate that Aerial-R1 significantly outperforms SFT baselines in segmentation precision while substantially mitigating fatal hallucinations in low-information scenarios, paving the way for more reliable and trustworthy aerial agents. Codes are available on https://github.com/phy709/Aerial-R1.

cs.CV↗

arXiv2026

A Complete Characterization of Learnability for Adversarial Noisy Bandits

We study adversarial noisy bandits given a known function class $\mathcal{F}$. In each round, the adversary selects a function $f \in \mathcal{F}$, the learner chooses an arm, and then observes a noisy reward determined by the chosen arm and the function $f$. The goal is to minimize the cumulative regret $R(T)$, defined as the difference between the learner's performance and that of the best fixed arm in hindsight over $T$ rounds. We say that a function class $\mathcal{F}$ is learnable if there exists an algorithm achieving sublinear regret. Our main result is a complete characterization of learnability for adversarial noisy bandits. The characterization is given in terms of a convexified variant of the generalized maximin volume introduced by Hanneke and Wang (2025): namely, the generalized maximin volume evaluated on the convex hull $\operatorname{co}(\mathcal F)$. We prove that $\mathcal F$ is learnable if and only if this convexified generalized maximin volume is positive at every scale. This condition characterizes learnability against both oblivious and adaptive adversaries, showing in particular that these two notions of learnability are equivalent in the noisy bandit setting. Our analysis reveals that the key complexity measure is closely connected to two new combinatorial notions, hitting set and distribution covering number, which may be of independent interest. These results establish the first complete characterization of learnability for adversarial noisy bandits.

cs.LG↗

arXiv2026

PINN-LPCDD: Physics-Informed Neural Network with Learnable and Physically Consistent Domain Decomposition for Compressible Multi-Medium Flows

The physics-informed neural network (PINN) faces significant challenges when solving compressible multi-medium flows, including the inability to locate interface positions, poor approximation of discontinuities, and difficulty in handling high-pressure-ratio problems between media with large stiffness differences. To address these issues, we propose PINN with Learnable and Physically Consistent Domain Decomposition (PINN-LPCDD), a framework that incorporates the sharp interface method (SIM) into PINNs for solving compressible multi-medium flows. Firstly, we design separate sub-networks for each medium with learnable parameters to automatically learn interface positions, so that the domain decomposition is physically consistent, and the contact discontinuity at the interface is naturally sharp. Secondly, the Euler equations with artificial viscosity are solved to enhance robustness when handling problems involving shock waves. Lastly, the scaling invariance of the Euler equations is exploited to facilitate the solution of high pressure ratio problems. Numerical experiments on air-air, air-helium, air-water, and air-R22 problems demonstrate the acceptable accuracy of the proposed method in simulating multi-medium flows.

physics.comp-ph↗

arXiv2026

FusionAudit: Pathway-Conditioned Robustness Auditing for Native Multimodal Models

As vision-language models (VLMs) transition from frozen-encoder pipelines to native multimodal architectures, the visual channel increasingly serves as both a perceptual input and a carrier of rendered text that competes with user prompts. Current single-attack, pixel-centric evaluations fail to capture this complexity: they collapse distinct visual attack surfaces into a single scalar and ignore how architectural shifts alter vulnerability. We introduce PaRMA, an evidence-based auditing framework that defines pathway-conditioned risk over two primary image-level interventions: pixel perturbation and rendered-text insertion. Instead of arbitrary attack enumeration, PaRMA uses pathway-specific diagnostics to expose architecture-specific blind spots. Notably, we uncover a see-but-not-deceived signature where native VLMs attend heavily to typographic overlays but visually reject them, and we identify an objective mismatch between encoder-level and answer-level attacks in deep-fusion models. To exploit these findings, we instantiate FusionAudit, introducing two diagnostic-motivated attacks: MANTIS (margin-aware pixel perturbation) and Nat-Evid (host-congruent semantic insertion). Evaluating 18 open-weight VLMs across a novel nativeness taxonomy, we demonstrate that legacy single-attack rankings (e.g., PGD) are unstable for native models. Our diagnostic-driven attacks substantially raise the estimated residual risk, showing that PaRMA provides a falsifiable, evidence-driven auditing methodology for uncovering vulnerabilities missed by standard benchmarks.

cs.CV↗

arXiv2026

Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory

Multimodal models and agents have emerged to enable AI systems to learn from and interact with the real world more like humans by integrating information across multiple modalities. However, enabling multimodal agents to continually acquire useful information from complex, heterogeneous multimodal data remains a significant challenge as they move toward long-horizon tasks and lifelong settings. Addressing this challenge requires effective memory mechanisms that can maintain continuity across sessions, adapt to changing users and environments, and refine memory-management policies through accumulated experience. We introduce a two-axis analytical framework for multimodal agent memory structured around two key questions: how multimodal information is formed and represented (Axis I, Formation and Representation), and how its management adapts with accumulated experience (Axis II, Memory Management). Axis I, Formation and Representation, characterizes what information is retained and how it is encoded (e.g., modality-specific embeddings, textual and structured records, latent and parametric representations) to support subsequent reasoning and action. Axis II, Memory Management, distinguishes three tiers: T1 Static Memory reuses fixed memory; T2 Dynamic Memory updates memory through a stable policy; and T3 Self-Evolving Memory updates memory through an evolving policy. Based on this framework, we synthesize recurring design patterns and derive principles for balancing source fidelity with abstraction, maintaining temporal validity and provenance, grounding memory use in recoverable evidence, and evolving memory policies over time. Furthermore, we discuss the challenges and potential future directions for the memory mechanism of multimodal agents.

cs.AI↗

arXiv2026

Quantum circuits for the advection-diffusion equation with boundary conditions based on LCHS

This paper proposes a systematic and explicit quantum circuit framework for solving advection-diffusion equations with boundary conditions, based on the Linear Combination of Hamiltonian Simulations (LCHS) method.By employing the Finite Volume Method (FVM) combined with various flux construction schemes, we elaborate the design of quantum circuits tailored explicitly for Robin boundary conditions (including Dirichlet and Neumann boundary conditions as special cases) and periodic boundary conditions. The Robin boundary elimination is encoded by newly constructed endpoint-projector circuits, and all physical and boundary-induced inhomogeneous terms are incorporated through an outer linear combination of unitaries (LCU). The error analysis adopts an LCU-weighted form, preserving the decay of the LCHS kernel in the product-formula bounds. Complementing this analysis, a comprehensive gate complexity analysis demonstrates the quantum advantages over classical computing in high-dimensional scenarios. We simulate the proposed circuits on a fault-tolerant emulator, and numerical results validate the effectiveness of the proposed framework across homogeneous, inhomogeneous, and high-dimensional cases.The proposed framework is compatible with numerous spatial discretization methods and numerical schemes, extends naturally to other linear PDEs, and establishes a practical foundation for solving large-scale PDE problems on future fault-tolerant quantum computers.

math.NA↗

arXiv2026

Remembering What Matters for Long-Horizon Robotic World Models

Autoregressive robot video world models based on discrete visual tokens remain unreliable over long rollout horizons. Retaining the entire generated history does not necessarily improve prediction quality. We investigate a possible memory-role conflict in Full Context decoding: persistent scene information, recent visual-action states, and stale generated history compete for attention despite their different predictive roles. This perspective motivates ReCAP (Receding Context with Anchor Preservation), a training-free context-scheduling method for long-horizon robot video world models. ReCAP keeps the pretrained tokenizer, world model, and visual decoder frozen. It preserves the initial scene anchor and a contiguous window of recent complete visual-action blocks while evicting older generated history. The schedule preserves within-block structure and temporal order, requires no additional modules or parameter updates, and bounds active-context growth independently of rollout length. We evaluate ReCAP on RT-1, BridgeV2, LIBERO-90, and CALVIN using frozen autoregressive world models. Across evaluated environments and horizons, ReCAP improves visual fidelity, action-conditioned consistency, and robot structure preservation relative to Full Context. On BridgeV2, it reduces LPIPS by 63.3%. In the system benchmark at H=64, it reduces the logical KV footprint by 73.3% and p50 latency by 6.51%. Attention analyses show increased attention to retained recent states, while history interventions associate larger amounts of stale context with higher prediction error. These observations are consistent with the memory-role-conflict hypothesis, although they do not isolate it from context-length and position-handling effects. Our results support inference-time context scheduling as a practical approach to improving long-horizon robot video prediction and computational efficiency without retraining.

cs.RO↗

arXiv2026

GroundSight at GroundLM 2026 Shared Tasks: GoldenViewVQA

GoldenViewVQA requires models to jointly answer driving-scene questions and identify the camera view containing the supporting visual evidence, making precise evidence localization as important as answer correctness. We present \textbf{CoVeR-VQA}, a training-free multi-stage verification and correction framework for grounded multi-view VQA. Starting from GPT-5.6 zero-shot predictions, CoVeR-VQA progressively applies view-specific verification with Gemini-3.6-Flash, prior-guided joint verification with Claude-Opus-5, and cross-split group-level verification that exploits semantically filtered question groups from shared multi-view scenes and validation-derived prior knowledge. On the official GoldenViewVQA test set, the four-stage CoVeR-VQA pipeline achieves 84.75\% Joint Accuracy, improving the GPT-5.6 zero-shot baseline by 13.56 percentage points, while reaching 94.92\% Answer Accuracy and 86.44\% View Accuracy. The final submitted run achieves 88.14\% Joint Accuracy after two additional evaluator-informed post-hoc corrections. Our analysis shows that supporting-view localization remains the primary source of residual errors, highlighting the importance of explicit evidence verification for reliable multi-view multimodal reasoning.

cs.CV↗

arXiv2026

Phase-aware video generation for physics-grounded dynamics and interactions

Generating physically plausible videos for solid-gas dynamics is challenging as different phases exhibit distinct dynamics yet remain coupled through physical interactions. We present PAVG, a Phase-Aware Video Generator for solid-gas dynamics and interactions. It employs a dual-branch architecture to explicitly model the distinct dynamics of solids and gases, while spatiotemporal cross-attention captures their physical interactions. This design enables PAVG to preserve phasespecific motion characteristics while producing physically consistent responses across phases. To facilitate this task, we further construct a simulation corpus comprising over 700K physical trajectories across diverse solid, gas, and solid-gas interaction scenarios. Extensive evaluations demonstrate that our PAVG produces videos with improved motion adherence, physical plausibility, and visual quality compared with existing approaches.

cs.CV↗

arXiv2026

SafeEvo: Deciphering the Safety Alignment Mechanism and Evolution in Language Models

Safety interpretability advances the study of Large Language Model (LLM) alignment from behavioral constraints driven by data or algorithms towards a deeper understanding of internal mechanisms. However, existing works have focused primarily on safety-related representations, attention heads, or neurons after alignment, while largely overlooking the safety mechanisms in pretrained-only models and their evolution across alignment checkpoints. To address this, we propose SafeEvo, an interpretability framework from the circuit (sparse subgraphs of an LLM) perspective. SafeEvo first applies an optimization-based extraction algorithm to identify weak refusal circuits in pretrained base LLMs that can independently express refusal behavior. Causally ablating these circuits completely eliminates the base model's refusal of harmful inputs. SafeEvo then traces the evolution of refusal circuits across successive alignment checkpoints and finds that their structures change progressively, suggesting that the alignment tax may result from refusal-circuit updates affecting utility-related parameters. To validate this, SafeEvo introduces Safety Circuit Alignment (SCA), which confines safety updates to the refusal circuits. Experiments across three LLMs and two alignment algorithms show that, on average, SCA outperforms vanilla alignment in three aspects: \textbf{(1) stronger alignment}, lowering harmfulness score by 63.21\%; \textbf{(2) less over-refusal}, yielding a 58.44\% decrease in refusal rates for benign queries; and \textbf{(3) better utility}, retaining 99.58\% of the original model capabilities.

cs.AI↗

arXiv2026

GeoAlign: Beyond Semantics with State-Guided Spatial Alignment in VLA Models

Current Vision-Language-Action (VLA) models often optimize for semantic grounding, whereas executable manipulation requires geometry-aware spatial alignment. We introduce GeoAlign, a state-guided spatial alignment architecture for VLA policy learning. Offline robot-domain RGB-D supervision post-trains an RGB geometry branch to produce Geometry-Enhanced Post-Trained (GEP) features; the depth head is then discarded, so policy training and rollout use RGB, language, and proprioceptive state. State-generated queries attend to the GEP grid to produce eight compact geometry tokens for action prediction. GeoAlign achieves 99.0% on LIBERO, 85.3% across three SimplerEnv-Fractal task families, and 78.8% on eight real-world ALOHA tasks, with ablations supporting the combined recipe of geometry post-training and state-guided querying on Isaac-GR00T N1.6-3B. Project website: https://chenyizhi123.github.io/geoalign-project/.

cs.RO↗