arXiv subjects

Chun Wang

Publications and source records attributed to Chun Wang.

At least 19 recordsLinked to original sources

arXiv2026

Simulating Cognitive Smart Freight Corridors with Agent-Based Models and Reinforcement Learning

Smart freight corridors offer a practical pathway for connected and automated vehicle (CAV) deployment in freight transportation, but physical experimentation is expensive and existing approaches rely on predefined control policies that cannot capture adaptive behaviors. This paper presents an agent-based modeling (ABM) framework coupling a physical infrastructure layer, a connectivity layer (V2X), and a decision layer integrating reinforcement learning (RL) and multi-agent reinforcement learning (MARL) for platoon formation and charging coordination. We evaluate three scenarios (Baseline, Assisted, and Cognitive) using throughput, congestion, energy, emissions, and robustness metrics. Preliminary results indicate that the Cognitive scenario achieves higher throughput and lower congestion than the baseline, while the Assisted scenario delivers meaningful energy savings per kilometer through platooning. Sensitivity analysis shows that the throughput advantage of the smart corridor widens under conditions with high demand and that MARL coordination extracts greater utilization from fixed charging capacity than rule-based assignment.

cs.ET↗

arXiv2026

CIDER: Continual Interactive Distillation for Embodied Reinforcement Learning

Human-in-the-loop real-world reinforcement learning enables rapid acquisition of effective robotic manipulation policies for individual tasks, often within tens of minutes. Yet it remains unclear how to extend this paradigm to continual learning, where a single policy must acquire new skills without losing previously learned behaviors. Existing real-world continual learning methods do not explicitly constrain prior behaviors, leading to severe catastrophic forgetting. We introduce Continual Interactive Distillation for Embodied Reinforcement Learning (CIDER), a continual reinforcement learning framework that freezes the accumulated historical policy as a teacher before learning each new task and interleaves task learning with distillation-based retention. We further introduce gradient routing to separate the gradients used for acquiring new tasks from those used for preserving prior behaviors. We evaluate our method with a single shared actor on six real-world household and industrial manipulation tasks. Interactive Distillation maintains high measured success on previously learned tasks across our six-task real-robot sequence while acquiring each new task in 10 to 20 minutes, whereas every baseline forgets at least one previous task. Additional ablations reveal the key design choices that govern the tradeoff between stability and plasticity in real-world continual reinforcement learning.

cs.RO↗

arXiv2026

Transient Depth Thermography for Probing Heat Transport

Directly probing heat propagation inside materials remains challenging because conventional measurements are predominantly sensitive to surface temperature. Depth thermography has enabled non-contact reconstruction of subsurface temperature profiles from spectrally resolved thermal radiation under steady-state conditions. Here, we extend this approach into the time domain, establishing transient depth thermography to resolve the evolution of internal temperature during heat transport. By exploiting wavelength-dependent optical penetration depth, time-resolved thermal-radiation spectra provide access to temperature as a function of both depth and time. Tracking this spatiotemporal temperature field enables direct probing of heat propagation and quantitative determination of out-of-plane thermal conductivity and interfacial thermal resistance. We demonstrate the approach in fused silica, obtaining thermal conductivity within 2% of established values, and measure the temperature-dependent thermal conductivity of MgF2 over a broad temperature range where existing data are sparse and inconsistent. Numerical simulations further demonstrate its extension to multilayer thin films for probing interfacial thermal resistance. By extending depth-resolved thermal spectroscopy from steady-state to transient heat transport, this work establishes a new optical route for non-contact characterization of thermal dynamics in bulk and layered materials.

physics.optics↗

arXiv2026

OPSD-V: On-Policy Self-Distillation for Post-Training Few-Step Autoregressive Video Generators

We propose OPSD-V, an on-policy self-distillation paradigm for post-training few-step autoregressive (AR) video diffusion models. Existing few-step AR video generators can produce long videos with low latency, but still suffer from error accumulation and weakened motion dynamics during long autoregressive rollout. OPSD-V reduces long-horizon degradation while preserving the original few-step inference path. The key idea is to introduce real long-video data as temporal context during training and use it to provide dense trajectory-level supervision. Specifically, the student follows the exact inference-time rollout, generating each chunk conditioned on its own previously generated KV cache. In parallel, the teacher is evaluated at the same student-visited denoising states, but uses a cleaner AR-consistent temporal cache in which older history can be replaced by real-video context. This provides dense denoising-level corrective targets under on-policy AR cache dynamics, without changing the sampler, number of denoising steps, or inference-time cache mechanism. We apply OPSD-V to representative few-step AR video models, including Self-Forcing and LongLive. Experiments show consistent improvements in visual quality, motion dynamics, and VBenchLong scores. A user study with 10 participants comparing 20 video pairs shows that OPSD-V is preferred over the base models in 66.0% of overall-preference judgments (82.5% excluding ties).

cs.CV↗

arXiv2026

RTE-FM-Dehazer: Radiative Transfer Equation Inspired Flow Matching for Real-World Image Dehazing

Single-image dehazing aims to recover a clear scene from a hazy image and is generally formulated as an image-to-image translation task; however, it faces two limitations. Its performance depends heavily on the haze-formation priors embedded in the model. Prevailing methods adopt the Atmospheric Scattering Model (ASM), whose assumptions of single scattering and homogeneous media are often violated, leading to residual haze and color drift. Moreover, large-scale real hazy/clear pairs are impractical to collect, and existing synthesis approaches fail to reproduce the full complexity of natural haze. To address these issues, we present RTE-FM-Dehazer, a novel dehazing approach, together with a scalable data pipeline. Unlike the ASM, the Radiative Transfer Equation (RTE) jointly accounts for both scattering and absorption, naturally accommodating the non-homogeneous, multiple-scattering media that characterize real hazy scenes. Motivated by the structural similarity between the RTE diffusion-absorption term and the ODE in flow matching, we introduce a diffusion-absorption regularizer derived from a reduced RTE, to steer the flow matching trajectory at each step. Next, leveraging modern vision-language models, we build an automated pipeline and release P-HAZE, a dataset of 50000 realistic hazy/clear pairs. Extensive evaluations demonstrate that RTE-FM-Dehazer, trained solely on P-HAZE, effectively eliminates artifacts like residual haze and color drift, exhibits strong cross-domain generalization, and achieves leading results on five real-world dehazing benchmarks.

cs.CV↗

arXiv2026

Multidimensional Item Response Theory under General Latent Distributions

Multidimensional item response theory (MIRT) provides an important psychometric framework for modeling how multiple latent traits jointly influence observed item responses. In most existing estimation procedures, the latent trait distribution is assumed to be Gaussian. Although computationally convenient, this assumption can be restrictive in many applications where the latent distribution exhibits skewness, heavy tails, or multimodality. More importantly, misspecifying the latent distribution may bias the estimation of item parameters and latent traits. To address this limitation, we propose a data-driven flow-based framework for MIRT models that can capture a broad class of non-Gaussian latent distributions. The proposed approach represents the latent distribution as an invertible transformation of a simple base distribution. For efficient estimation, we further introduce a conditional flow as a function of both the observed response and the noise to approximate the posterior distribution. Under this framework, the item parameters, latent distribution, and posterior approximation can be learned jointly. Comprehensive simulation studies show that the proposed method improves item-parameter and latent-trait recovery when the true latent distribution is non-normal. An application to a personality dataset further illustrates the practical utility of the proposed framework for modeling complex latent trait distributions in large-scale data.

stat.ME↗

arXiv2026

Enhancing Zero-shot Personalized Image Aesthetics Assessment with Profile-aware Multimodal LLM

Personalized image aesthetics assessment (PIAA) aims to predict an individual user's subjective rating of an image, which requires modeling user-specific aesthetic preferences. Existing methods rely on historical user ratings for this modeling and therefore struggle when such data are unavailable. We address this zero-shot setting by using user profiles as contextual signals for personalization and adopting a profile-based personalization paradigm. We introduce P-MLLM, a profile-aware multimodal LLM that augments a frozen LLM with selective fusion modules for controlled visual integration. These modules selectively integrate visual information into the model's evolving hidden states during profile-conditioned reasoning, allowing visual information to be incorporated in a profile-aware manner. Experiments on recent PIAA benchmarks show that P-MLLM achieves competitive zero-shot performance and remains effective even with coarse profile information, highlighting the potential of profile-based personalization for zero-shot PIAA.

cs.CV↗

arXiv2026

Proof of the Andrews-El Bachraoui positivity conjecture

We prove that for $k\ge 1$, all coefficients in the expansion of the series $$\sum_{n\ge 0} \frac{(q^{2n+2}, q^{2n+2k}; q^2)_\infty}{(q^{2n+1};q^2)_\infty^2} q^{2n}$$ are positive, by $q$-hypergeometric means. This confirms a recent conjecture of Andrews and El Bachraoui.

math.CO↗

arXiv2025

Source Localization and Power Estimation through RISs: Performance Analysis and Prototype Validations

This paper investigates the capabilities and effectiveness of backward localization centered on reconfigurable intelligent surfaces (RISs). In the backward sensing paradigm, the region of interest (RoI) is illuminated using a set of diverse radiation patterns. These patterns encode spatial information into a sequence of measurements, which are subsequently processed to reconstruct the RoI. We show that a single RIS can estimate the direction of arrival of incident waves by leveraging configurational diversity, and that the spatial diversity provided by multiple RISs further improves the accuracy of source localization and power estimation. The underlying structure of the sensing operator in the multi-snapshot measurement process is clarified. For single-RIS localization, the sensing operator is decomposed into a product of structured matrices, each corresponding to a specific physical process: wave propagation to and from the RIS, the relative phase offsets of elements with respect to the reference point, and the applied phase configuration of each element. A unified framework for identifying key performance indicators is established by analyzing the conditioning of the sensing operators. In the multi-RIS setting, we derive--via rank analysis--the governing law among the RoI size, the number of elements, and the number of measurements. Upper bounds on the relative error of the least squares reconstruction algorithm are derived. These bounds clarify how key performance indicators affect estimation error and provide valuable guidance for system-level optimization. Numerical experiments confirm that the trend of the relative error is consistent with the theoretical bounds.

eess.SP↗

arXiv2025

Supervised Fine Tuning of Large Language Models for Domain Specific Knowledge Graph Construction:A Case Study on Hunan's Historical Celebrities

Large language models and knowledge graphs offer strong potential for advancing research on historical culture by supporting the extraction, analysis, and interpretation of cultural heritage. Using Hunan's modern historical celebrities shaped by Huxiang culture as a case study, pre-trained large models can help researchers efficiently extract key information, including biographical attributes, life events, and social relationships, from textual sources and construct structured knowledge graphs. However, systematic data resources for Hunan's historical celebrities remain limited, and general-purpose models often underperform in domain knowledge extraction and structured output generation in such low-resource settings. To address these issues, this study proposes a supervised fine-tuning approach for enhancing domain-specific information extraction. First, we design a fine-grained, schema-guided instruction template tailored to the Hunan historical celebrities domain and build an instruction-tuning dataset to mitigate the lack of domain-specific training corpora. Second, we apply parameter-efficient instruction fine-tuning to four publicly available large language models - Qwen2.5-7B, Qwen3-8B, DeepSeek-R1-Distill-Qwen-7B, and Llama-3.1-8B-Instruct - and develop evaluation criteria for assessing their extraction performance. Experimental results show that all models exhibit substantial performance gains after fine-tuning. Among them, Qwen3-8B achieves the strongest results, reaching a score of 89.3866 with 100 samples and 50 training iterations. This study provides new insights into fine-tuning vertical large language models for regional historical and cultural domains and highlights their potential for cost-effective applications in cultural heritage knowledge extraction and knowledge graph construction.

cs.CL↗

arXiv2025

PANER: A Paraphrase-Augmented Framework for Low-Resource Named Entity Recognition

Named Entity Recognition (NER) is a critical task that requires substantial annotated data, making it challenging in low-resource scenarios where label acquisition is expensive. While zero-shot and instruction-tuned approaches have made progress, they often fail to generalize to domain-specific entities and do not effectively utilize limited available data. We present a lightweight few-shot NER framework that addresses these challenges through two key innovations: (1) a new instruction tuning template with a simplified output format that combines principles from prior IT approaches to leverage the large context window of recent state-of-the-art LLMs; (2) introducing a strategic data augmentation technique that preserves entity information while paraphrasing the surrounding context, thereby expanding our training data without compromising semantic relationships. Experiments on benchmark datasets show that our method achieves performance comparable to state-of-the-art models on few-shot and zero-shot tasks, with our few-shot approach attaining an average F1 score of 80.1 on the CrossNER datasets. Models trained with our paraphrasing approach show consistent improvements in F1 scores of up to 17 points over baseline versions, offering a promising solution for groups with limited NER training data and compute power.

cs.CL↗

arXiv2025

Bilateral Two-Parameter Mock Theta Functions and Related Applications

In this paper, we investigate new relationships for bilateral series related to two-parameter mock theta functions, which lead to many identities concerning the bilateral mock theta functions. In addition, interesting relations between the classical mock theta functions and the bilateral series are also concluded.

math.NT↗

arXiv2025

Fluid Language Model Benchmarking

Language model (LM) benchmarking faces several challenges: comprehensive evaluations are costly, benchmarks often fail to measure the intended capabilities, and evaluation quality can degrade due to labeling errors and benchmark saturation. Although various strategies have been proposed to mitigate these issues, they tend to address individual aspects in isolation, neglecting broader questions about overall evaluation quality. Here, we introduce Fluid Benchmarking, a new evaluation approach that advances LM benchmarking across multiple dimensions. Inspired by psychometrics, Fluid Benchmarking is based on the insight that the relative value of benchmark items depends on an LM's capability level, suggesting that evaluation should adapt to each LM. Methodologically, Fluid Benchmarking estimates an item response model based on existing LM evaluation results and uses the inferred quantities to select evaluation items dynamically, similar to computerized adaptive testing in education. In our experiments, we compare Fluid Benchmarking against the common practice of random item sampling as well as more sophisticated baselines, including alternative methods grounded in item response theory. We examine four dimensions -- efficiency, validity, variance, and saturation -- and find that Fluid Benchmarking achieves superior performance in all of them (e.g., higher validity and less variance on MMLU with fifty times fewer items). Our analysis shows that the two components of Fluid Benchmarking have distinct effects: item response theory, used to map performance into a latent ability space, increases validity, while dynamic item selection reduces variance. Overall, our results suggest that LM benchmarking can be substantially improved by moving beyond static evaluation.

cs.CL↗

arXiv2025

A novel biomass fluidized bed gasification model coupled with machine learning and CFD simulation

A coupling model of biomass fluidized bed gasification based on machine learning and computational fluid dynamics is proposed to improve the prediction accuracy and computational efficiency of complex thermochemical reaction process. By constructing a high-quality data set based on experimental data and high fidelity simulation results, the agent model used to describe the characteristics of reaction kinetics was trained and embedded into the computational fluid dynamics (CFD) framework to realize the real-time update of reaction rate and composition evolution.

cs.LG↗

arXiv2025

Revisiting the Radial Metallicity Gradient-Age Relation in the Milky Way's Thin and Thick Disks

Galactic disks typically exhibit a negative radial metallicity gradient, indicating faster enrichment in the inner regions. Recent studies report that this gradient becomes flatter with increasing stellar age in the Milky Way's (MW) thin disk, while the thick disk exhibits a mildly positive gradient across all ages. In this work, we revisit the metallicity gradient-age relation (MGAR) in both the thin and thick disks of the MW, using spectroscopic data from LAMOST DR8 and stellar ages calibrated with asteroseismology. Our results show a steadily flattening MGAR in the thin disk and confirm a positive gradient $\sim0.013\,\mathrm{dex\,kpc^{-1}}$ in the thick disk. The flattening in the thin disk may be caused by large-scale radial migration induced by transient spiral arms, or by a time-dependent steepening of the interstellar medium (ISM) metallicity gradient as suggested by recent FIRE2 simulations. The positive gradient in the thick disk may reflect early enrichment of the outer regions by strong feedback or starburst-driven outflows in a turbulent, gas-rich proto-disk. These findings suggest distinct chemodynamical evolution paths for the MW's thin and thick disks and provide valuable constraints for future models of Galactic chemical evolution.

astro-ph.GA↗

arXiv2025

Dual Information Speech Language Models for Emotional Conversations

Conversational systems relying on text-based large language models (LLMs) often overlook paralinguistic cues, essential for understanding emotions and intentions. Speech-language models (SLMs), which use speech as input, are emerging as a promising solution. However, SLMs built by extending frozen LLMs struggle to capture paralinguistic information and exhibit reduced context understanding. We identify entangled information and improper training strategies as key issues. To address these issues, we propose two heterogeneous adapters and suggest a weakly supervised training strategy. Our approach disentangles paralinguistic and linguistic information, enabling SLMs to interpret speech through structured representations. It also preserves contextual understanding by avoiding the generation of task-specific vectors through controlled randomness. This approach trains only the adapters on common datasets, ensuring parameter and data efficiency. Experiments demonstrate competitive performance in emotional conversation tasks, showcasing the model's ability to effectively integrate both paralinguistic and linguistic information within contextual settings.

cs.CL↗

arXiv2025

Optimal Transport Regularization for Speech Text Alignment in Spoken Language Models

Spoken Language Models (SLMs), which extend Large Language Models (LLMs) to perceive speech inputs, have gained increasing attention for their potential to advance speech understanding tasks. However, despite recent progress, studies show that SLMs often struggle to generalize across datasets, even for trained languages and tasks, raising concerns about whether they process speech in a text-like manner as intended. A key challenge underlying this limitation is the modality gap between speech and text representations. The high variability in speech embeddings may allow SLMs to achieve strong in-domain performance by exploiting unintended speech variations, ultimately hindering generalization. To mitigate this modality gap, we introduce Optimal Transport Regularization (OTReg), a method that formulates speech-text alignment as an optimal transport problem and derives a regularization loss to improve SLM training. In each training iteration, OTReg first establishes a structured correspondence between speech and transcript embeddings by determining the optimal transport plan, then incorporates the regularization loss based on this transport plan to optimize SLMs in generating speech embeddings that align more effectively with transcript embeddings. OTReg is lightweight, requiring no additional labels or learnable parameters, and integrates seamlessly into existing SLM training procedures. Extensive multilingual ASR experiments demonstrate that OTReg enhances speech-text alignment, mitigates the modality gap, and consequently improves SLM generalization across diverse datasets.

cs.CL↗

arXiv2025

A North-South Metallicity Asymmetry in the Outer Galactic disk -- Evidence for the Pericentric Passage of the Sagittarius Dwarf Galaxy

We present maps of the mean metallicity distributions on the Galactocentric $R$--$Z$ plane at different azimuthal angles using red clump stars selected from the LAMOST and APOGEE surveys. In the inner disk ($R < $ 11\,kpc), the metallicity distribution is symmetric between the upper and lower disk. However, we find a North-South metallicity asymmetry in the outer disk ($R > 11$\,kpc), especially towards the anti-Galactic center ($-5^\circ < \Phi < 15^\circ$) direction. By further dissecting the map in age space, we detect this asymmetry across all mono-age stellar populations. However, the asymmetry is less pronounced in older populations ($\tau > 8$ Gyr) compared to younger ones ($\tau < 6$\,Gyr). This reduced significance likely stems from three factors: larger age uncertainties, fewer stars in the outer disk, and the kinematically hotter nature of older populations. The observed metallicity asymmetry may be the consequence of the purturbation of the recent pericentric passage through the Galactic disk and tidal force of the well-known Sagittarius dwarf galaxy.

astro-ph.GA↗