arXiv subjects

Guang Lu

Publications and source records attributed to Guang Lu.

5 recordsLinked to original sources

arXiv2026

Type-III solar radio bursts with spike-like toppings

Spike-type III burst pairs represent a distinct class of solar radio emissions in which clusters of spike-like bursts appear atop the highfrequency onset of type III bursts. Using high time-frequency resolution data from the Chashan Broadband Solar radio spectrometer at meter wavelengths (CBSm), we present the largest statistical study to date of such events, comprising 502 spike-type III pairs from 35 events recorded between November 2023 and October 2025. We find that spike-like clusters systematically precede their associated type III bursts by 0.5-3 s in time (~87% of pairs) and by 3-30 MHz in frequency (~80%), a temporal and spectral offset that differs from earlier reports. The spike-like clusters exhibit diverse morphologies, including point-like, blob-like, drifting, and diffuse structures, with durations of ~0.5-5 s and bandwidths of 15-150 MHz. Bi-directional drifting structures with rates of ~20-100 MHz s-1 are observed, consistent with source motion both toward and away from the Sun. Furthermore, spike emission is predominantly strongly circularly polarized, with more than 64% of clusters showing maximum polarization exceeding 0.6, in stark contrast to the generally weak polarization of type III bursts. These findings point to an origin of the spike radiation in a multiscale, inhomogeneous, and highly dynamic electron-acceleration region, providing novel observational constraints on the mechanisms underlying coherent solar radio bursts.

astro-ph.SR↗

arXiv2023

A Contextualized Real-Time Multimodal Emotion Recognition for Conversational Agents using Graph Convolutional Networks in Reinforcement Learning

Owing to the recent developments in Generative Artificial Intelligence (GenAI) and Large Language Models (LLM), conversational agents are becoming increasingly popular and accepted. They provide a human touch by interacting in ways familiar to us and by providing support as virtual companions. Therefore, it is important to understand the user's emotions in order to respond considerately. Compared to the standard problem of emotion recognition, conversational agents face an additional constraint in that recognition must be real-time. Studies on model architectures using audio, visual, and textual modalities have mainly focused on emotion classification using full video sequences that do not provide online features. In this work, we present a novel paradigm for contextualized Emotion Recognition using Graph Convolutional Network with Reinforcement Learning (conER-GRL). Conversations are partitioned into smaller groups of utterances for effective extraction of contextual information. The system uses Gated Recurrent Units (GRU) to extract multimodal features from these groups of utterances. More importantly, Graph Convolutional Networks (GCN) and Reinforcement Learning (RL) agents are cascade trained to capture the complex dependencies of emotion features in interactive scenarios. Comparing the results of the conER-GRL model with other state-of-the-art models on the benchmark dataset IEMOCAP demonstrates the advantageous capabilities of the conER-GRL architecture in recognizing emotions in real-time from multimodal conversational signals.

cs.CL↗

arXiv2023

Hybrid Long Document Summarization using C2F-FAR and ChatGPT: A Practical Study

Text summarization is a downstream natural language processing (NLP) task that challenges the understanding and generation capabilities of language models. Considerable progress has been made in automatically summarizing short texts, such as news articles, often leading to satisfactory results. However, summarizing long documents remains a major challenge. This is due to the complex contextual information in the text and the lack of open-source benchmarking datasets and evaluation frameworks that can be used to develop and test model performance. In this work, we use ChatGPT, the latest breakthrough in the field of large language models (LLMs), together with the extractive summarization model C2F-FAR (Coarse-to-Fine Facet-Aware Ranking) to propose a hybrid extraction and summarization pipeline for long documents such as business articles and books. We work with the world-renowned company getAbstract AG and leverage their expertise and experience in professional book summarization. A practical study has shown that machine-generated summaries can perform at least as well as human-written summaries when evaluated using current automated evaluation metrics. However, a closer examination of the texts generated by ChatGPT through human evaluations has shown that there are still critical issues in terms of text coherence, faithfulness, and style. Overall, our results show that the use of ChatGPT is a very promising but not yet mature approach for summarizing long documents and can at best serve as an inspiration for human editors. We anticipate that our work will inform NLP researchers about the extent to which ChatGPT's capabilities for summarizing long documents overlap with practitioners' needs. Further work is needed to test the proposed hybrid summarization pipeline, in particular involving GPT-4, and to propose a new evaluation framework tailored to the task of summarizing long documents.

cs.CL↗

arXiv2018

Carbonate-Bridged Dinuclear and Trinuclear Dysprosium(III) Single-Molecule Magnets

In 2016, we reported a single-ion magnet [Dy(bbpen)Br] with an energy barrier over 1000 K. Here a dimeric [Dy2(mu-CO3)(bbpen)2(H2O)].H2O.CH3OH (1) and a trimeric [Dy3(mu3-CO3)(bppen)3](CF3SO3).H2O (2) single-molecule magnets (SMMs) were obtained through replacing the Br- anion with the carbonate bridge. Their effective relaxation barriers at zero dc field are decrease to 51 K and 422 K, respectively, which are consist with their structural modifications.

cond-mat.mtrl-sci↗

arXiv2016

Enhanced absorption in heterostructures with graphene and truncated photonic crystals

We theoretically and experimentally investigate the optical absorption properties of heterostructures composed of graphene films and truncated photonic crystals (PCs) in the visible range. The experimental results show that the absorption of the heterostructure is enhanced greatly in the whole forbidden gap of PCs compared with that of graphene alone. The absorption is enhanced over a wide angle of incidence for both transverse electric (TE) and transverse magnetic (TM) polarizations. The enhanced absorption band broadens for TE polarization but narrows for TM polarization as the incident angle increases. In the forbidden gap of the PCs, the maximum absorptance of the heterostructures is nearly four times of that of bare graphene films for arbitrary incident angles and polarizations. The optical experiments are in excellent agreement with the theoretical results.

physics.optics↗