arXiv subjects

Jing Li

Publications and source records attributed to Jing Li.

At least 325 records · Page 18Linked to original sources

arXiv2025

Rocketquake Seismology with a Falcon 9 Rocket Source

We investigate the feasibility of using rocket launches, specifically rocketquakes, as a seismic source to image subsurface velocity and geology of planetary bodies. Toward this goal, we record the seismic vibrations excited by a Falcon 9 rocket launch from Vandenberg Space Force Base (SFB) near Lompoc, California. Nine passive three-component (3C) seismometers were deployed every 18.75 meters along a 45-degree line from the launch site starting at the offset of about 7 km kilometers from the launch pad. Results show that coherent body waves can be recorded with a P-velocity of more than 2.0 km/s and a penetration depth of 1 km or deeper. Stronger Rayleigh waves were also recorded and inverted to give an S-velocity profile to a depth of 60 m. Notably, a 3C recorder placed approximately 15 km from the launch site did not capture any discernible body wave arrivals. The imaging techniques employed for rocketquake seismology integrate inversion methods from earthquake and exploration seismology, yielding the P- and S-velocity profiles of the subsurface. These results suggest that rocket launches as seismic sources will provide unprecedented opportunities for identifying the subsurface hazards, faults, tunnels, water ice, and mineral deposits of planetary bodies and their moons.

physics.geo-ph↗

arXiv2025

Concept-Aware Latent and Explicit Knowledge Integration for Enhanced Cognitive Diagnosis

Cognitive diagnosis can infer the students' mastery of specific knowledge concepts based on historical response logs. However, the existing cognitive diagnostic models (CDMs) represent students' proficiency via a unidimensional perspective, which can't assess the students' mastery on each knowledge concept comprehensively. Moreover, the Q-matrix binarizes the relationship between exercises and knowledge concepts, and it can't represent the latent relationship between exercises and knowledge concepts. Especially, when the granularity of knowledge attributes refines increasingly, the Q-matrix becomes incomplete correspondingly and the sparse binary representation (0/1) fails to capture the intricate relationships among knowledge concepts. To address these issues, we propose a Concept-aware Latent and Explicit Knowledge Integration model for cognitive diagnosis (CLEKI-CD). Specifically, a multidimensional vector is constructed according to the students' mastery and exercise difficulty for each knowledge concept from multiple perspectives, which enhances the representation capabilities of the model. Moreover, a latent Q-matrix is generated by our proposed attention-based knowledge aggregation method, and it can uncover the coverage degree of exercises over latent knowledge. The latent Q-matrix can supplement the sparse explicit Q-matrix with the inherent relationships among knowledge concepts, and mitigate the knowledge coverage problem. Furthermore, we employ a combined cognitive diagnosis layer to integrate both latent and explicit knowledge, further enhancing cognitive diagnosis performance. Extensive experiments on real-world datasets demonstrate that CLEKI-CD outperforms the state-of-the-art models. The proposed CLEKI-CD is promising in practical applications in the field of intelligent education, as it exhibits good interpretability with diagnostic results.

cs.LG↗

arXiv2025

Cross-Modal Synergies: Unveiling the Potential of Motion-Aware Fusion Networks in Handling Dynamic and Static ReID Scenarios

Navigating the complexities of person re-identification (ReID) in varied surveillance scenarios, particularly when occlusions occur, poses significant challenges. We introduce an innovative Motion-Aware Fusion (MOTAR-FUSE) network that utilizes motion cues derived from static imagery to significantly enhance ReID capabilities. This network incorporates a dual-input visual adapter capable of processing both images and videos, thereby facilitating more effective feature extraction. A unique aspect of our approach is the integration of a motion consistency task, which empowers the motion-aware transformer to adeptly capture the dynamics of human motion. This technique substantially improves the recognition of features in scenarios where occlusions are prevalent, thereby advancing the ReID process. Our comprehensive evaluations across multiple ReID benchmarks, including holistic, occluded, and video-based scenarios, demonstrate that our MOTAR-FUSE network achieves superior performance compared to existing approaches.

cs.CV↗

arXiv2025

Generative Visual Commonsense Answering and Explaining with Generative Scene Graph Constructing

Visual Commonsense Reasoning, which is regarded as one challenging task to pursue advanced visual scene comprehension, has been used to diagnose the reasoning ability of AI systems. However, reliable reasoning requires a good grasp of the scene's details. Existing work fails to effectively exploit the real-world object relationship information present within the scene, and instead overly relies on knowledge from training memory. Based on these observations, we propose a novel scene-graph-enhanced visual commonsense reasoning generation method named \textit{\textbf{G2}}, which first utilizes the image patches and LLMs to construct a location-free scene graph, and then answer and explain based on the scene graph's information. We also propose automatic scene graph filtering and selection strategies to absorb valuable scene graph information during training. Extensive experiments are conducted on the tasks and datasets of scene graph constructing and visual commonsense answering and explaining, respectively. Experimental results and ablation analysis demonstrate the effectiveness of our proposed framework.

cs.CV↗

arXiv2025

FLM-101B: An Open LLM and How to Train It with $100K Budget

Large language models (LLMs) are considered important approaches towards foundational machine intelligence, achieving remarkable success in Natural Language Processing and multimodal tasks, among others. However, the carbon footprints and financial costs originating from heavy pre-training computation is a non-negligible issue. Progressive training methods, inspired by the neurogenesis process that grows neural structures, have shown potential to accelerate LLM pre-training. However, the algorithms, implementation, and practices for progressively training LLMs beyond 100B parameters remain underexplored. In this paper, we show that our model, namely FLM-101B, trained with our growth strategy under a budget of \$100K, reaches 80\% of the baselines' performances with only 10\% of their floating-point operations. We believe that further studies on progressive training will benefit the community by cutting down the costs and promoting green AI. The checkpoint of FLM-101B is released at https://huggingface.co/CofeAI/FLM-101B.

cs.CL↗

arXiv2025

Dual-Modality Representation Learning for Molecular Property Prediction

Molecular property prediction has attracted substantial attention recently. Accurate prediction of drug properties relies heavily on effective molecular representations. The structures of chemical compounds are commonly represented as graphs or SMILES sequences. Recent advances in learning drug properties commonly employ Graph Neural Networks (GNNs) based on the graph representation. For the SMILES representation, Transformer-based architectures have been adopted by treating each SMILES string as a sequence of tokens. Because each representation has its own advantages and disadvantages, combining both representations in learning drug properties is a promising direction. We propose a method named Dual-Modality Cross-Attention (DMCA) that can effectively combine the strengths of two representations by employing the cross-attention mechanism. DMCA was evaluated across eight datasets including both classification and regression tasks. Results show that our method achieves the best overall performance, highlighting its effectiveness in leveraging the complementary information from both graph and SMILES modalities.

cs.LG↗

arXiv2025

Human-In-the-Loop Software Development Agents

Recently, Large Language Models (LLMs)-based multi-agent paradigms for software engineering are introduced to automatically resolve software development tasks (e.g., from a given issue to source code). However, existing work is evaluated based on historical benchmark datasets, rarely considers human feedback at each stage of the automated software development process, and has not been deployed in practice. In this paper, we introduce a Human-in-the-loop LLM-based Agents framework (HULA) for software development that allows software engineers to refine and guide LLMs when generating coding plans and source code for a given task. We design, implement, and deploy the HULA framework into Atlassian JIRA for internal uses. Through a multi-stage evaluation of the HULA framework, Atlassian software engineers perceive that HULA can minimize the overall development time and effort, especially in initiating a coding plan and writing code for straightforward tasks. On the other hand, challenges around code quality remain a concern in some cases. We draw lessons learned and discuss opportunities for future work, which will pave the way for the advancement of LLM-based agents in software development.

cs.SE↗

arXiv2025

Supertetragonal BaSnO3 induced giant ferroelectricity in SrTiO3/BaSnO3 superlattices

Perovskite BaSnO3 has an Sn s-orbital conduction band minimum, which makes it of interest as a transparent-conducting oxide parent compound but also contraindicates the ferroelectric instability characteristic of the related compound BaTiO3. In this work, we studied the effect of (001) compressive strain on BaSnO3 using first-principles methods. We found that, with low compressive strain, symmetry breaking takes cubic BaSnO3 to a nonpolar tetragonal state, with a first-order phase transition to a hidden highly-polarized ferroelectric supertetragonal state at about -5%. Based on the facts that the mismatch of lattice constant in experiment between BaSnO3 and SrTiO3 is about -5.2% and coherent growth of BaSnO3 on SrTiO3 has been experimentally realized for BaSnO3 layers thinner than 3 unit-cells, we studied a series of SrTiO3/BaSnO3 superlattices with one or two unit-cells of BaSnO3 and several unit-cells of SrTiO3. We found that the superlattices are ferroelectric with large polarizations. We propose that the origin of ferroelectricity in the superlattices is the mechanical and electrical coupling of the BaSnO3 and SrTiO3 layers, with polarized supertetragonal state of BaSnO3 induced by compressive-strain from the SrTiO3 layers and polarization of the SrTiO3 layers by the polar BaSnO3 layers. Due to the distinctive electronic states in the BaSnO3 layers, the realization of ferroelectricity holds promise for the design of novel electronic devices.

cond-mat.mtrl-sci↗

arXiv2025

Debate-to-Write: A Persona-Driven Multi-Agent Framework for Diverse Argument Generation

Writing persuasive arguments is a challenging task for both humans and machines. It entails incorporating high-level beliefs from various perspectives on the topic, along with deliberate reasoning and planning to construct a coherent narrative. Current language models often generate surface tokens autoregressively, lacking explicit integration of these underlying controls, resulting in limited output diversity and coherence. In this work, we propose a persona-based multi-agent framework for argument writing. Inspired by the human debate, we first assign each agent a persona representing its high-level beliefs from a unique perspective, and then design an agent interaction process so that the agents can collaboratively debate and discuss the idea to form an overall plan for argument writing. Such debate process enables fluid and nonlinear development of ideas. We evaluate our framework on argumentative essay writing. The results show that our framework can generate more diverse and persuasive arguments through both automatic and human evaluations.

cs.CL↗

arXiv2024

AnglE-optimized Text Embeddings

High-quality text embedding is pivotal in improving semantic textual similarity (STS) tasks, which are crucial components in Large Language Model (LLM) applications. However, a common challenge existing text embedding models face is the problem of vanishing gradients, primarily due to their reliance on the cosine function in the optimization objective, which has saturation zones. To address this issue, this paper proposes a novel angle-optimized text embedding model called AnglE. The core idea of AnglE is to introduce angle optimization in a complex space. This novel approach effectively mitigates the adverse effects of the saturation zone in the cosine function, which can impede gradient and hinder optimization processes. To set up a comprehensive STS evaluation, we experimented on existing short-text STS datasets and a newly collected long-text STS dataset from GitHub Issues. Furthermore, we examine domain-specific STS scenarios with limited labeled data and explore how AnglE works with LLM-annotated data. Extensive experiments were conducted on various tasks including short-text STS, long-text STS, and domain-specific STS tasks. The results show that AnglE outperforms the state-of-the-art (SOTA) STS models that ignore the cosine saturation zone. These findings demonstrate the ability of AnglE to generate high-quality text embeddings and the usefulness of angle optimization in STS.

cs.CL↗

arXiv2024

Unlocking adaptive digital pathology through dynamic feature learning

Foundation models have revolutionized the paradigm of digital pathology, as they leverage general-purpose features to emulate real-world pathological practices, enabling the quantitative analysis of critical histological patterns and the dissection of cancer-specific signals. However, these static general features constrain the flexibility and pathological relevance in the ever-evolving needs of clinical applications, hindering the broad use of the current models. Here we introduce PathFiT, a dynamic feature learning method that can be effortlessly plugged into various pathology foundation models to unlock their adaptability. Meanwhile, PathFiT performs seamless implementation across diverse pathology applications regardless of downstream specificity. To validate PathFiT, we construct a digital pathology benchmark with over 20 terabytes of Internet and real-world data comprising 28 H\&E-stained tasks and 7 specialized imaging tasks including Masson's Trichrome staining and immunofluorescence images. By applying PathFiT to the representative pathology foundation models, we demonstrate state-of-the-art performance on 34 out of 35 tasks, with significant improvements on 23 tasks and outperforming by 10.20% on specialized imaging tasks. The superior performance and versatility of PathFiT open up new avenues in computational pathology.

eess.IV↗

arXiv2024

Knowledge Editing with Dynamic Knowledge Graphs for Multi-Hop Question Answering

Multi-hop question answering (MHQA) poses a significant challenge for large language models (LLMs) due to the extensive knowledge demands involved. Knowledge editing, which aims to precisely modify the LLMs to incorporate specific knowledge without negatively impacting other unrelated knowledge, offers a potential solution for addressing MHQA challenges with LLMs. However, current solutions struggle to effectively resolve issues of knowledge conflicts. Most parameter-preserving editing methods are hindered by inaccurate retrieval and overlook secondary editing issues, which can introduce noise into the reasoning process of LLMs. In this paper, we introduce KEDKG, a novel knowledge editing method that leverages a dynamic knowledge graph for MHQA, designed to ensure the reliability of answers. KEDKG involves two primary steps: dynamic knowledge graph construction and knowledge graph augmented generation. Initially, KEDKG autonomously constructs a dynamic knowledge graph to store revised information while resolving potential knowledge conflicts. Subsequently, it employs a fine-grained retrieval strategy coupled with an entity and relation detector to enhance the accuracy of graph retrieval for LLM generation. Experimental results on benchmarks show that KEDKG surpasses previous state-of-the-art models, delivering more accurate and reliable answers in environments with dynamic information.

cs.CL↗

arXiv2024

Tackling the Dynamicity in a Production LLM Serving System with SOTA Optimizations via Hybrid Prefill/Decode/Verify Scheduling on Efficient Meta-kernels

Meeting growing demands for low latency and cost efficiency in production-grade large language model (LLM) serving systems requires integrating advanced optimization techniques. However, dynamic and unpredictable input-output lengths of LLM, compounded by these optimizations, exacerbate the issues of workload variability, making it difficult to maintain high efficiency on AI accelerators, especially DSAs with tile-based programming models. To address this challenge, we introduce XY-Serve, a versatile, Ascend native, end-to-end production LLM-serving system. The core idea is an abstraction mechanism that smooths out the workload variability by decomposing computations into unified, hardware-friendly, fine-grained meta primitives. For attention, we propose a meta-kernel that computes the basic pattern of matmul-softmax-matmul with architectural-aware tile sizes. For GEMM, we introduce a virtual padding scheme that adapts to dynamic shape changes while using highly efficient GEMM primitives with assorted fixed tile sizes. XY-Serve sits harmoniously with vLLM. Experimental results show up to 89% end-to-end throughput improvement compared with current publicly available baselines on Ascend NPUs. Additionally, our approach outperforms existing GEMM (average 14.6% faster) and attention (average 21.5% faster) kernels relative to existing libraries. While the work is Ascend native, we believe the approach can be readily applicable to SIMT architectures as well.

cs.AI↗

arXiv2024

Demise of Kreutz Sungrazing Comet C/2024 S1 (ATLAS)

Most Kreutz family sungrazing comets are discovered only days before perihelion, severely limiting observational opportunities to study their physical nature and decay. Kreutz sungrazer C/2024 S1 (ATLAS) was discovered a month before reaching its perihelion distance of 0.008 au, allowing physical observations from both ground- and space-based telescopes. We present observations from 0.9 au to 0.4 au using the Nordic Optical Telescope showing that 1) nucleus disintegration was on-going already at 0.7 au pre-perihelion, 2) the activity varied unpredictably with distance and 3) the nucleus radius was $<$600 m (red geometric albedo 0.04 assumed). We also use coronagraphic observations from the STEREO-A spacecraft to study C/2024 S1 at heliocentric distances $\lesssim$0.1 au. We find that the coma scattering cross-section peaked near 0.075 au and faded progressively, by a factor $\sim$20, towards the last observation at 0.02 au. We interpret the near-perihelion fading as a result of the sublimation of refractory coma grains, beginning at blackbody temperatures $\sim$1000 K, consistent with olivine composition. The comet was not detected after perihelion. We consider processes operating to destroy the nucleus when near perihelion, concluding that rotational instability and sublimation losses work together towards this end, even before entry of the comet into the Roche lobe of the Sun.

astro-ph.EP↗

arXiv2024

Strongly interacting matter in extreme magnetic fields

Magnetic fields are ubiquitous across different physical systems of current interest; from the early Universe, compact astrophysical objects and heavy-ion collisions to condensed matter systems. A proper treatment of the effects produced by magnetic fields during the dynamical evolution of these systems, can help to understand observables that otherwise show a puzzling behavior. Furthermore, when these fields are comparable to or stronger than Λ_QCD, they serve as excellent probes to help elucidate the physics of strongly interacting matter under extreme conditions of temperature and density. In this work we provide a comprehensive review of recent developments on the description of QED and QCD systems where magnetic field driven effects are important. These include the modification of meson static properties such as masses and form factors, the chiral magnetic effect, the description of anomalous transport coefficients, superconductivity in extreme magnetic fields, the properties of neutron stars, the evolution of heavy-ion collisions, as well as effects on the QCD phase diagram. We describe recent theory and phenomenological developments using effective models as well as LQCD methods. The work represents a state-of-the-art review of the field, motivated by presentations and discussions during the "Workshop on Strongly Interacting Matter in Strong Electromagnetic Fields" that took place in the European Centre for Theoretical Studies in Nuclear Physics and Related Areas (ECT*) in the city of Trento, Italy, September 25-29, 2023.

nucl-th↗

arXiv2024

UniPoll: A Unified Social Media Poll Generation Framework via Multi-Objective Optimization

Social media platforms are vital for expressing opinions and understanding public sentiment, yet many analytical tools overlook passive users who mainly consume content without engaging actively. To address this, we introduce UniPoll, an advanced framework designed to automatically generate polls from social media posts using sophisticated natural language generation (NLG) techniques. Unlike traditional methods that struggle with social media's informal and context-sensitive nature, UniPoll leverages enriched contexts from user comments and employs multi-objective optimization to enhance poll relevance and engagement. To tackle the inherently noisy nature of social media data, UniPoll incorporates Retrieval-Augmented Generation (RAG) and synthetic data generation, ensuring robust performance across real-world scenarios. The framework surpasses existing models, including T5, ChatGLM3, and GPT-3.5, in generating coherent and contextually appropriate question-answer pairs. Evaluated on the Chinese WeiboPolls dataset and the newly introduced English RedditPolls dataset, UniPoll demonstrates superior cross-lingual and cross-platform capabilities, making it a potent tool to boost user engagement and create a more inclusive environment for interaction.

cs.CL↗

arXiv2024

Impromptu Cybercrime Euphemism Detection

Detecting euphemisms is essential for content security on various social media platforms, but existing methods designed for detecting euphemisms are ineffective in impromptu euphemisms. In this work, we make a first attempt to an exploration of impromptu euphemism detection and introduce the Impromptu Cybercrime Euphemisms Detection (ICED) dataset. Moreover, we propose a detection framework tailored to this problem, which employs context augmentation modeling and multi-round iterative training. Our detection framework mainly consists of a coarse-grained and a fine-grained classification model. The coarse-grained classification model removes most of the harmless content in the corpus to be detected. The fine-grained model, impromptu euphemisms detector, integrates context augmentation and multi-round iterations training to better predicts the actual meaning of a masked token. In addition, we leverage ChatGPT to evaluate the mode's capability. Experimental results demonstrate that our approach achieves a remarkable 76-fold improvement compared to the previous state-of-the-art euphemism detector.

cs.CL↗

arXiv2024

Experimental electronic phase diagram in a diamond-lattice antiferromagnetic system

We report Ni-doping effect on the magnetic and electronic properties of thiospinel Co$_{1-x}$Ni$_x$[Co$_{0.3}$Ir$_{1.7}$]S$_4$ (0 $\leq x \leq$ 1). The parent compound Co[Co$_{0.3}$Ir$_{1.7}$]S$_4$ exhibits antiferromagnetic order below $T_\mathrm{N} \sim$ 292 K within the $A$-site diamond sublattice, along with a narrow charge-transfer gap. Upon Ni doping, an insulator-to-metal crossover occurs at $x \sim$ 0.35, and the antiferromagnetism is gradually suppressed, with $T_\mathrm{N}$ decreasing to 23 K at $x =$ 0.7. In the metallic state, a spin-glass-like transition emerges at low temperatures. The antiferromagnetic transition is completely suppressed at $x_\mathrm{c} \sim$ 0.95, around which a non-Fermi-liquid behavior emerges, evident from the $T^α$ temperature dependence with $α\approx$ 1.2-1.3 in resistivity and divergent behavior of $C/T$ in specific heat at low temperatures. Meanwhile, the electronic specific heat coefficient $γ$ increases substantially, signifying an enhancement of the quasiparticle effective mass. The magnetic phase diagram has been established, in which an antiferromagnetic quantum critical point is avoided at $x_\mathrm{c}$. Conversely, the observed glass-like tail above the critical concentration aligns more closely with theoretical predictions for an extended region of quantum Griffiths phase in the presence of strong disorder.

cond-mat.str-el↗