arXiv subjects

Yu Rong

Publications and source records attributed to Yu Rong.

At least 73 records · Page 4Linked to original sources

arXiv2025

FINEREASON: Evaluating and Improving LLMs' Deliberate Reasoning through Reflective Puzzle Solving

Many challenging reasoning tasks require not just rapid, intuitive responses, but a more deliberate, multi-step approach. Recent progress in large language models (LLMs) highlights an important shift from the "System 1" way of quick reactions to the "System 2" style of reflection-and-correction problem solving. However, current benchmarks heavily rely on the final-answer accuracy, leaving much of a model's intermediate reasoning steps unexamined. This fails to assess the model's ability to reflect and rectify mistakes within the reasoning process. To bridge this gap, we introduce FINEREASON, a logic-puzzle benchmark for fine-grained evaluation of LLMs' reasoning capabilities. Each puzzle can be decomposed into atomic steps, making it ideal for rigorous validation of intermediate correctness. Building on this, we introduce two tasks: state checking, and state transition, for a comprehensive evaluation of how models assess the current situation and plan the next move. To support broader research, we also provide a puzzle training set aimed at enhancing performance on general mathematical tasks. We show that models trained on our state checking and transition data demonstrate gains in math reasoning by up to 5.1% on GSM8K.

cs.CL↗

arXiv2025

A Survey of Geometric Graph Neural Networks: Data Structures, Models and Applications

Geometric graphs are a special kind of graph with geometric features, which are vital to model many scientific problems. Unlike generic graphs, geometric graphs often exhibit physical symmetries of translations, rotations, and reflections, making them ineffectively processed by current Graph Neural Networks (GNNs). To address this issue, researchers proposed a variety of geometric GNNs equipped with invariant/equivariant properties to better characterize the geometry and topology of geometric graphs. Given the current progress in this field, it is imperative to conduct a comprehensive survey of data structures, models, and applications related to geometric GNNs. In this paper, based on the necessary but concise mathematical preliminaries, we formalize geometric graph as the data structure, on top of which we provide a unified view of existing models from the geometric message passing perspective. Additionally, we summarize the applications as well as the related datasets to facilitate later research for methodology development and experimental evaluation. We also discuss the challenges and future potential directions of geometric GNNs at the end of this survey.

cs.LG↗

arXiv2025

Disentangling Language and Culture for Evaluating Multilingual Large Language Models

This paper introduces a Dual Evaluation Framework to comprehensively assess the multilingual capabilities of LLMs. By decomposing the evaluation along the dimensions of linguistic medium and cultural context, this framework enables a nuanced analysis of LLMs' ability to process questions within both native and cross-cultural contexts cross-lingually. Extensive evaluations are conducted on a wide range of models, revealing a notable "CulturalLinguistic Synergy" phenomenon, where models exhibit better performance when questions are culturally aligned with the language. This phenomenon is further explored through interpretability probing, which shows that a higher proportion of specific neurons are activated in a language's cultural context. This activation proportion could serve as a potential indicator for evaluating multilingual performance during model training. Our findings challenge the prevailing notion that LLMs, primarily trained on English data, perform uniformly across languages and highlight the necessity of culturally and linguistically model evaluations. Our code can be found at https://yingjiahao14. github.io/Dual-Evaluation/.

cs.CL↗

arXiv2025

Materials Generation in the Era of Artificial Intelligence: A Comprehensive Survey

Materials are the foundation of modern society, underpinning advancements in energy, electronics, healthcare, transportation, and infrastructure. The ability to discover and design new materials with tailored properties is critical to solving some of the most pressing global challenges. In recent years, the growing availability of high-quality materials data combined with rapid advances in Artificial Intelligence (AI) has opened new opportunities for accelerating materials discovery. Data-driven generative models provide a powerful tool for materials design by directly create novel materials that satisfy predefined property requirements. Despite the proliferation of related work, there remains a notable lack of up-to-date and systematic surveys in this area. To fill this gap, this paper provides a comprehensive overview of recent progress in AI-driven materials generation. We first organize various types of materials and illustrate multiple representations of crystalline materials. We then provide a detailed summary and taxonomy of current AI-driven materials generation approaches. Furthermore, we discuss the common evaluation metrics and summarize open-source codes and benchmark datasets. Finally, we conclude with potential future directions and challenges in this fast-growing field. The related sources can be found at https://github.com/ZhixunLEE/Awesome-AI-for-Materials-Generation.

cond-mat.mtrl-sci↗

arXiv2025

Flow Along the K-Amplitude for Generative Modeling

In this work, we propose a novel generative learning paradigm, K-Flow, an algorithm that flows along the $K$-amplitude. Here, $k$ is a scaling parameter that organizes frequency bands (or projected coefficients), and amplitude describes the norm of such projected coefficients. By incorporating the $K$-amplitude decomposition, K-Flow enables flow matching across the scaling parameter as time. We discuss three venues and six properties of K-Flow, from theoretical foundations, energy and temporal dynamics, and practical applications, respectively. Specifically, from the practical usage perspective, K-Flow allows steerable generation by controlling the information at different scales. To demonstrate the effectiveness of K-Flow, we conduct experiments on unconditional image generation, class-conditional image generation, and molecule assembly generation. Additionally, we conduct three ablation studies to demonstrate how K-Flow steers scaling parameter to effectively control the resolution of image generation.

cs.LG↗

arXiv2025

A negative stellar mass$-$gaseous metallicity gradient relation of dwarf galaxies modulated by stellar feedback

Baryonic cycling is reflected in the spatial distribution of metallicity within galaxies, yet gas-phase metallicity distribution and its connection with other properties of dwarf galaxies are largely unexplored. We present the first systematic study of radial gradients of gas-phase metallicities for a sample of 55 normal nearby star-forming dwarf galaxies (stellar mass $M_\star$ ranging from $10^7$ to $10^{9.5}\ M_\odot$), based on MUSE spectroscopic observations. We find that metallicity gradient shows a significant negative correlation (correlation coefficient $r \approx -0.56$) with $\log M_\star$, in contrast to the flat or even positive correlation observed for higher-mass galaxies. This negative correlation is accompanied by a stronger central suppression of metallicity compared to the outskirts in lower-mass galaxies. Among the other explored galaxy properties-including baryonic mass, star formation distribution, galaxy environment, regularity of the gaseous velocity field, and effective yield of metals $y_{\rm eff}$-only the velocity field regularity and $y_{\rm eff}$ show residual correlation with the metallicity gradient after controlling for $M_\star$, in the sense that galaxies with irregular velocity fields or lower $y_{\rm eff}$ tend to have less negative or more positive gradients. Particularly, a linear combination of $\log M_\star$ and $\log y_{\rm eff}$ significantly improves the correlation with metallicity gradient ($r \approx -0.68$) compared to $\log M_\star$ alone. The lack of correlation with environment disfavors gas accretion as a dominant factor. Our findings imply that metal mixing and transport processes, including but not limited to feedback-driven outflows, are more important than in-situ metal production in shaping the metallicity distribution of dwarf galaxies.

astro-ph.GA↗

arXiv2025

Unexpected clustering pattern in dwarf galaxies challenges formation models

The galaxy correlation function serves as a fundamental tool for studying cosmology, galaxy formation, and the nature of dark matter. It is well established that more massive, redder and more compact galaxies tend to have stronger clustering in space. These results can be understood in terms of galaxy formation in Cold Dark Matter (CDM) halos of different mass and assembly history. Here, we report an unexpectedly strong large-scale clustering for isolated, diffuse and blue dwarf galaxies, comparable to that seen for massive galaxy groups but much stronger than that expected from their halo mass. Our analysis indicates that the strong clustering aligns with the halo assembly bias seen in simulations with the standard $Λ$CDM cosmology only if more diffuse dwarfs formed in low-mass halos of older ages. This pattern is not reproduced by existing models of galaxy evolution in a $Λ$CDM framework, and our finding provides new clues for the search of more viable models. Our results can be explained well by assuming self-interacting dark matter, suggesting that such a scenario should be considered seriously.

astro-ph.CO↗

arXiv2025

Alignment of Blue/Green and Red Early-type Galaxies with Large-scale Filaments Reveals Distinct Evolutionary Pathways

We investigate the alignment of non-red early-type galaxies (ETGs) with blue or green colors within large-scale filaments and compare this alignment pattern with that of red ETGs. Our analysis reveals a significant alignment of the major axes of red ETGs with the orientations of their host cosmic filaments, consistent with prior research. In contrast, non-red ETGs show no significant alignment signal. This divergence in alignment behavior between non-red and red ETGs plausibly suggests distinct evolutionary pathways for non-red and red ETGs.

astro-ph.GA↗

arXiv2025

A Speech Production Model for Radar: Connecting Speech Acoustics with Radar-Measured Vibrations

Millimeter Wave (mmWave) radar has emerged as a promising modality for speech sensing, offering advantages over traditional microphones. Prior works have demonstrated that radar captures motion signals related to vocal vibrations, but there is a gap in the understanding of the analytical connection between radar-measured vibrations and acoustic speech signals. We establish a mathematical framework linking radar-captured neck vibrations to speech acoustics. We derive an analytical relationship between neck surface displacements and speech. We use data from 66 human participants, and statistical spectral distance analysis to empirically assess the model. Our results show that the radar-measured signal aligns more closely with our model filtered vibration signal derived from speech than with raw speech itself. These findings provide a foundation for improved radar-based speech processing for applications in speech enhancement, coding, surveillance, and authentication.

eess.AS↗

arXiv2025

LUCAS: Layered Universal Codec Avatars

Photorealistic 3D head avatar reconstruction faces critical challenges in modeling dynamic face-hair interactions and achieving cross-identity generalization, particularly during expressions and head movements. We present LUCAS, a novel Universal Prior Model (UPM) for codec avatar modeling that disentangles face and hair through a layered representation. Unlike previous UPMs that treat hair as an integral part of the head, our approach separates the modeling of the hairless head and hair into distinct branches. LUCAS is the first to introduce a mesh-based UPM, facilitating real-time rendering on devices. Our layered representation also improves the anchor geometry for precise and visually appealing Gaussian renderings. Experimental results indicate that LUCAS outperforms existing single-mesh and Gaussian-based avatar models in both quantitative and qualitative assessments, including evaluations on held-out subjects in zero-shot driving scenarios. LUCAS demonstrates superior dynamic performance in managing head pose changes, expression transfer, and hairstyle variations, thereby advancing the state-of-the-art in 3D head avatar reconstruction.

cs.CV↗

arXiv2025

The baryonic Tully-Fisher relation of HI-bearing low-surface brightness galaxies implies their formation mechanism

We investigate the baryonic Tully-Fisher relation in low surface brightness galaxies selected from the Arecibo Legacy Fast ALFA survey. We find that the $\rm HI$-bearing low surface brightness galaxies still follow the baryonic Tully-Fisher relation of typical late-type galaxies, with a slope of approximately 4 in the baryonic mass versus rotational velocity diagram on the logarithmic scale, i.e., $M_{\rm{b}}\propto v_{\rm{rot}}^4$. Our findings suggest that the matter distributions in low surface brightness galaxies may resemble that of general late-type galaxies, and hint that low surface brightness galaxies may not originate from dark matter halos of low densities or stronger/weaker feedback processes, but may emerge from dark matter halos with high spin values.

astro-ph.GA↗

arXiv2025

Orthogonal Alignment of Galaxy Group Angular Momentum with Cosmic Filament Spines: An Observational Study

We investigate the alignment between the angular momenta of galaxy groups and the spines of their associated cosmic filaments. Our results demonstrate a significant tendency for these two orientations to be perpendicular, indicating that the rotation of a galaxy group does not originate from the spin of cosmic filaments. Instead, it is driven by the orbital angular momentum contributed by member galaxies as they accrete along the direction of the filament spines. Moreover, the strength of this perpendicular alignment signal varies with the richness of the galaxy groups, with the most pronounced alignment observed among the wealthiest groups. This pronounced alignment is largely due to the more coherent spatial distribution of member galaxies in richer groups relative to the filament spines. Our study provides valuable insights into the mechanisms of angular momentum acquisition in galaxy groups from an observational standpoint.

astro-ph.GA↗

arXiv2025

Adaptive Extensive Cancellation Algorithm and Harmonic Enhanced Heart Rate Estimation based on MMWave Radar

Heart rate (HR) monitoring is crucial for assessing physical fitness, cardiovascular health, and stress management. Millimeter-wave radar offers a promising noncontact solution for long-term monitoring. However, accurate HR estimation remains challenging in low signal-tonoise ratio (SNR) conditions. To deal with both respiration harmonics and intermodulation interference, this paper proposes a cancellation-before-estimation strategy. Firstly, we present the adaptive extensive cancellation algorithm (ECA) to suppress respiratory and its low-order harmonics. Then, we propose an adaptive harmonic enhanced trace (AHET) method to avoid intermodulation interference by refining the HR search region. Various experimental results validate the effectiveness of the proposed methods, demonstrating improvements in accuracy, robustness, and computational efficiency compared to conventional approaches based on the FMCW (Frequency Modulated Continuous Wave) system

eess.SP↗

arXiv2025

Halo Spin Dependence on Environment for HI-bearing galaxies

Leveraging the semi-analytic method, we compute halo spins for a substantial sample of HI-bearing galaxies observed in the Arecibo Legacy Fast Alfa Survey. Our statistical analysis reveals a correlation between halo spin and environment, although the trend is subtle. On average, galaxies exhibit a decreasing halo spin tendency in denser environments. This observation contrasts with previous results from $N$-body simulations in the Lambda cold dark matter framework. The discrepancy may be attributed to environmental gas stripping, leading to an underestimation of halo spins in galaxies in denser environments, or to baryonic processes that significantly alter the original dark matter halo spins, deviating from previous $N$-body simulation findings.

astro-ph.GA↗

arXiv2025

Atomas: Hierarchical Alignment on Molecule-Text for Unified Molecule Understanding and Generation

Molecule-and-text cross-modal representation learning has emerged as a promising direction for enhancing the quality of molecular representation, thereby improving performance in various scientific fields. However, most approaches employ a global alignment approach to learn the knowledge from different modalities that may fail to capture fine-grained information, such as molecule-and-text fragments and stereoisomeric nuances, which is crucial for downstream tasks. Furthermore, it is incapable of modeling such information using a similar global alignment strategy due to the lack of annotations about the fine-grained fragments in the existing dataset. In this paper, we propose Atomas, a hierarchical molecular representation learning framework that jointly learns representations from SMILES strings and text. We design a Hierarchical Adaptive Alignment model to automatically learn the fine-grained fragment correspondence between two modalities and align these representations at three semantic levels. Atomas's end-to-end training framework supports understanding and generating molecules, enabling a wider range of downstream tasks. Atomas achieves superior performance across 12 tasks on 11 datasets, outperforming 11 baseline models thus highlighting the effectiveness and versatility of our method. Scaling experiments further demonstrate Atomas's robustness and scalability. Moreover, visualization and qualitative analysis, validated by human experts, confirm the chemical relevance of our approach. Codes are released on https://github.com/yikunpku/Atomas.

q-bio.QM↗

arXiv2025

InversionGNN: A Dual Path Network for Multi-Property Molecular Optimization

Exploring chemical space to find novel molecules that simultaneously satisfy multiple properties is crucial in drug discovery. However, existing methods often struggle with trading off multiple properties due to the conflicting or correlated nature of chemical properties. To tackle this issue, we introduce InversionGNN framework, an effective yet sample-efficient dual-path graph neural network (GNN) for multi-objective drug discovery. In the direct prediction path of InversionGNN, we train the model for multi-property prediction to acquire knowledge of the optimal combination of functional groups. Then the learned chemical knowledge helps the inversion generation path to generate molecules with required properties. In order to decode the complex knowledge of multiple properties in the inversion path, we propose a gradient-based Pareto search method to balance conflicting properties and generate Pareto optimal molecules. Additionally, InversionGNN is able to search the full Pareto front approximately in discrete chemical space. Comprehensive experimental evaluations show that InversionGNN is both effective and sample-efficient in various discrete multi-objective settings including drug discovery.

cs.LG↗

arXiv2025

Babel: Open Multilingual Large Language Models Serving Over 90% of Global Speakers

Large language models (LLMs) have revolutionized natural language processing (NLP), yet open-source multilingual LLMs remain scarce, with existing models often limited in language coverage. Such models typically prioritize well-resourced languages, while widely spoken but under-resourced languages are often overlooked. To address this disparity, we introduce $\texttt{Babel}$, an open multilingual LLM that covers the top 25 languages by number of speakers, supports over 90% of the global population, and includes many languages neglected by other open multilingual LLMs. Unlike traditional continue pretraining approaches, Babel expands its parameter count through a layer extension technique that elevates Babel's performance ceiling. We introduce two variants: $\texttt{Babel-9B}$, designed for efficient inference and fine-tuning, and $\texttt{Babel-83B}$, which sets a new standard for open multilingual LLMs. Extensive evaluations on multilingual tasks demonstrate its superior performance compared to open LLMs of comparable size. In addition, using open-source supervised fine-tuning datasets, Babel achieves remarkable performance, with Babel-9B-Chat leading among 10B-sized LLMs and Babel-83B-Chat setting a new standard for multilingual tasks, reaching the same level of commercial models.

cs.CL↗

arXiv2025

CirT: Global Subseasonal-to-Seasonal Forecasting with Geometry-inspired Transformer

Accurate Subseasonal-to-Seasonal (S2S) climate forecasting is pivotal for decision-making including agriculture planning and disaster preparedness but is known to be challenging due to its chaotic nature. Although recent data-driven models have shown promising results, their performance is limited by inadequate consideration of geometric inductive biases. Usually, they treat the spherical weather data as planar images, resulting in an inaccurate representation of locations and spatial relations. In this work, we propose the geometric-inspired Circular Transformer (CirT) to model the cyclic characteristic of the graticule, consisting of two key designs: (1) Decomposing the weather data by latitude into circular patches that serve as input tokens to the Transformer; (2) Leveraging Fourier transform in self-attention to capture the global information and model the spatial periodicity. Extensive experiments on the Earth Reanalysis 5 (ERA5) reanalysis dataset demonstrate our model yields a significant improvement over the advanced data-driven models, including PanguWeather and GraphCast, as well as skillful ECMWF systems. Additionally, we empirically show the effectiveness of our model designs and high-quality prediction over spatial and temporal dimensions.

cs.LG↗