arXiv subjects

Yao Ma

Publications and source records attributed to Yao Ma.

At least 37 records · Page 2Linked to original sources

arXiv2026

A Wall Function for Turbulent Boundary Layers under Rotation via Symbolic Regression

This study employs symbolic regression to derive physically interpretable, white-box wall-function expressions for turbulent boundary layers under system rotation. Flows in a rotating frame are subject to Coriolis forces, which deflect the boundary layer profile from static case. The classical law of the wall, formulated under non-rotating conditions, is ill-suited to describing the effects of rotation. To obtain the wall function under rotation, we examine the deflection behavior of the turbulent boundary layers on the leading and trailing sides, and construct wall functions that are valid over a wide range of rotation numbers. The analytical expressions show that, as the rotation effect intensifies, the boundary layer on the leading side contracts whereas that on the trailing side expands, and the leading side exhibits a tendency towards relaminarization, consistent with high-fidelity numerical results. The resulting symbolic expressions are compact and interpretable. The wall functions obtained in this study complement conventional wall functions, and provide a new avenue for turbulence model closure subject to system rotation.

physics.flu-dyn↗

arXiv2026

Unifying and Optimizing Data Values for Selection via Sequential Decision-Making

Data selection has emerged as a crucial downstream application of data valuation, yet the theoretical foundations for using data values in selection remain underexplored. We reformulate data selection as a sequential decision-making problem where the optimal selection sequence arises from dynamic programming, and data values can be understood as encodings of this optimal sequence. This framework unifies and reinterprets existing methods like Data Shapley through the lens of approximate dynamic programming, revealing them as myopic linear approximations to the sequential problem. We further analyze how selection optimality degrades with utility curvature under submodularity, explaining when and why these approximations fail. To bridge theory and practice, we propose an efficient bipartite graph-based surrogate that preserves submodular structure while enabling scalable greedy selection with provable guarantees. Experiments on classical ML benchmarks and large-scale LLM fine-tuning data selection demonstrate substantial improvements over existing methods. Code is publicly available at https://github.com/frankhlchi/SeqDataVal

cs.AI↗

arXiv2026

ARES: Adaptive Red-Teaming and End-to-End Repair of Policy-Reward System

Reinforcement Learning from Human Feedback (RLHF) is central to aligning Large Language Models (LLMs), yet it introduces a critical vulnerability: an imperfect Reward Model (RM) can become a single point of failure when it fails to penalize unsafe behaviors. While existing red-teaming approaches primarily target policy-level weaknesses, they overlook what we term systemic weaknesses cases where both the core LLM and the RM fail in tandem. We present ARES, a framework that systematically discovers and mitigates such dual vulnerabilities. ARES employs a ``Safety Mentor'' that dynamically composes semantically coherent adversarial prompts by combining structured component types (topics, personas, tactics, goals) and generates corresponding malicious and safe responses. This dual-targeting approach exposes weaknesses in both the core LLM and the RM simultaneously. Using the vulnerabilities gained, ARES implements a two-stage repair process: first fine-tuning the RM to better detect harmful content, then leveraging the improved RM to optimize the core model. Experiments across multiple adversarial safety benchmarks demonstrate that ARES substantially enhances safety robustness while preserving model capabilities, establishing a new paradigm for comprehensive RLHF safety alignment.

cs.AI↗

arXiv2026

Nuclear matter properties and neutron star structures from an extended linear sigma model

The properties of nuclear matter and the structures of neutron stars are analyzed with a baryonic extended linear sigma model in mean-field approximation, where the masses of baryons and mesons are generated via the spontaneous chiral symmetry breaking. The couplings between the iso-scalar scalar meson and nucleons, $g_{σNN}$, the iso-vector scalar meson and nucleons, $g_{a_0 NN}$, and the four-vector meson couplings play an important role in the properties of nuclear matter and neutron stars. The introduction of the $δ$ meson leads to a plateau structure of the symmetry energy, $E_{\rm sym}(n)$, at intermediate densities, which is crucial to the consistency of neutron skin thickness of $^{208}$Pb and the tidal deformability of a canonical neutron star. The explicit chiral symmetry breaking term is then introduced with a constant background field, $ξ$, which can be related to the current quark mass and thus the pion-nucleon sigma term, $σ_{πN}$. A negative $σ_{πN}$ leads to a stiffer EOS of neutron star matter and thus a larger maximum mass of neutron stars, but the value of $σ_{πN}$ needed to satisfy the astrophysical constraints is negative, not positive as the vacuum value. The study may provide insights into the running behaviors of the parameters in the low-energy effective model to give the density-dependent description for the EOS of neutron star matter.

nucl-th↗

arXiv2026

Understanding Structured Financial Data with LLMs: A Case Study on Fraud Detection

Detecting fraud in financial transactions typically relies on tabular models that demand heavy feature engineering to handle high-dimensional data and offer limited interpretability, making it difficult for humans to understand predictions. Large Language Models (LLMs), in contrast, can produce human-readable explanations and facilitate feature analysis, potentially reducing the manual workload of fraud analysts and informing system refinements. However, they perform poorly when applied directly to tabular fraud detection due to the difficulty of reasoning over many features, the extreme class imbalance, and the absence of contextual information. To bridge this gap, we introduce FinFRE-RAG, a two-stage approach that applies importance-guided feature reduction to serialize a compact subset of numeric/categorical attributes into natural language and performs retrieval-augmented in-context learning over label-aware, instance-level exemplars. Across four public fraud datasets and three families of open-weight LLMs, FinFRE-RAG substantially improves F1/MCC over direct prompting and is competitive with strong tabular baselines in several settings. Although these LLMs still lag behind specialized classifiers, they narrow the performance gap and provide interpretable rationales, highlighting their value as assistive tools in fraud analysis.

cs.LG↗

arXiv2026

Neutron star structure and nuclear matter properties from a general Walecka-type model with Bayesian analysis

We establish a Bayesian analysis framework with a general Walecka-type relativistic mean-field model to study dense nuclear matter under constraints from nuclear matter properties and neutron star observations. With experimental and observational data well described, we find that pure hadronic descriptions can generate a peak structure in sound velocity by $ω$, $ρ$, $σ$, and $a_0$ meson mixing, which is crucial for describing both medium and massive neutron stars. As the peak structure is frequently interpreted as a signature of phase transitions, our findings provide a new perspective on the microscopic origin of the sound velocity peak just with pure hadronic matter.

nucl-th↗

arXiv2026

DPDisc: From Factoid Questions to Data Product Requests for Open-World Data Product Discovery over Tables and Text

Data products are reusable, self-contained assets designed for specific business use cases. Automating their discovery is of great industry interest, as it enables efficient data access in large data lakes and supports analytical workflows. However, no benchmark currently exists for data product discovery over hybrid table-text corpora. Existing datasets focus on answering single factoid questions over individual tables rather than assembling multiple related data assets into coherent products. To address this gap, we present DPDisc, the first large-scale benchmark for data product discovery, where systems must retrieve coherent collections of tables and passages to satisfy high-level Data Product Requests (DPRs). We introduce DPForge, an automated pipeline that systematically repurposes table-text QA datasets by clustering related tables and passages into coherent data products, generating professional-level analytical requests using an LLM ensemble, and validating quality through multi-phase LLM evaluation. DPDisc comprises 13,076 validated instances with full provenance, derived from three representative datasets spanning open-domain and financial domains. Baseline experiments with sparse, dense, and hybrid retrieval methods imply evaluation feasibility while revealing substantial performance gaps across domains, indicating opportunities for future research in structure-aware data product discovery. Code and datasets are available at: Dataset: https://huggingface.co/datasets/ibm-research/data-product-benchmark Code: https://github.com/ibm/data-product-benchmark

cs.IR↗

arXiv2026

Chiral-scale effective field theory for dense and thermal systems

We established a new power counting scheme, chiral-scale density counting (CSDC) rules, for the application of the chiral-scale effective field theory to nuclear matter at finite densities and temperatures. Within this framework, the free fermion gas is at the leading order, while one-boson-exchange interactions appear at the next-to-leading order, and the multi-meson couplings are at higher orders. Then, we applied the CSDC rules to study the nuclear matter properties, and estimated the valid regions of the CSDC rules. It was found that the zero temperature symmetric nuclear matter properties around saturation density and the critical temperature of liquid-gas phase transition can be captured by an appropriate choice of CSDC orders, and the results beyond these regions are align with the chiral nuclear force. Moreover, the evolution of scale symmetry was found to be consistent with previous studies. The results of this work indicate that the quantum corrections may be crucial in the studies of nuclear matter in a wide density region.

nucl-th↗

arXiv2026

A Practical Post-Quantum Distributed Ledger Protocol for Financial Institutions

Traditional financial institutions face inefficiencies that can be addressed by distributed ledger technology. However, a primary barrier to adoption is the privacy concerns surrounding publicly available transaction data. Existing private protocols for distributed ledger that focus on the Ring-CT model are not suitable for adoption for financial institutions. We propose a post-quantum, lattice-based transaction scheme for encrypted ledgers which better aligns with institutions' requirements for confidentiality and audit-ability. The construction leverages various zero-knowledge proof techniques, and introduces a new method for equating two commitment messages, without the capability to open one of the commitment during the re-commitment. Subsequently, we build a publicly verifiable transaction scheme that is efficient for single or multi-assets, by introducing a new compact range-proof. We then provide a security analysis of it. The techniques used and the proofs constructed could be of independent interest.

cs.CR↗

arXiv2026

Adversarial Déjà Vu: Jailbreak Dictionary Learning for Stronger Generalization to Unseen Attacks

Large language models remain vulnerable to jailbreak attacks that bypass safety guardrails to elicit harmful outputs. Defending against novel jailbreaks represents a critical challenge in AI safety. Adversarial training -- designed to make models robust against worst-case perturbations -- has been the dominant paradigm for adversarial robustness. However, due to optimization challenges and difficulties in defining realistic threat models, adversarial training methods often fail on newly developed jailbreaks in practice. This paper proposes a new paradigm for improving robustness against unseen jailbreaks, centered on the Adversarial Déjà Vu hypothesis: novel jailbreaks are not fundamentally new, but largely recombinations of adversarial skills from previous attacks. We study this hypothesis through a large-scale analysis of 32 attack papers published over two years. Using an automated pipeline, we extract and compress adversarial skills into a sparse dictionary of primitives, with LLMs generating human-readable descriptions. Our analysis reveals that unseen attacks can be effectively explained as sparse compositions of earlier skills, with explanatory power increasing monotonically as skill coverage grows. Guided by this insight, we introduce Adversarial Skill Compositional Training (ASCoT), which trains on diverse compositions of skill primitives rather than isolated attack instances. ASCoT substantially improves robustness to unseen attacks, including multi-turn jailbreaks, while maintaining low over-refusal rates. We also demonstrate that expanding adversarial skill coverage, not just data scale, is key to defending against novel attacks. \textcolor{red}{\textbf{Warning: This paper contains content that may be harmful or offensive in nature.

cs.LG↗

arXiv2026

SaVe-TAG: LLM-based Interpolation for Long-Tailed Text-Attributed Graphs

Real-world graph data often follows long-tailed distributions, making it difficult for Graph Neural Networks (GNNs) to generalize well across both head and tail classes. Recent advances in Vicinal Risk Minimization (VRM) have shown promise in mitigating class imbalance with numeric interpolation; however, existing approaches largely rely on embedding-space arithmetic, which fails to capture the rich semantics inherent in text-attributed graphs. In this work, we propose our method, SaVe-TAG (Semantic-aware Vicinal Risk Minimization for Long-Tailed Text-Attributed Graphs), a novel VRM framework that leverages Large Language Models (LLMs) to perform text-level interpolation, generating on-manifold, boundary-enriching synthetic samples for minority classes. To mitigate the risk of noisy generation, we introduce a confidence-based edge assignment mechanism that uses graph topology as a natural filter to ensure structural consistency. We provide theoretical justification for our method and conduct extensive experiments on benchmark datasets, showing that our approach consistently outperforms both numeric interpolation and prior long-tailed node classification baselines. Our results highlight the importance of integrating semantic and structural signals for balanced and effective learning on text-attributed graphs. The source code is publicly available at: https://github.com/LWang-Laura/SaVe-TAG.

cs.AI↗

arXiv2026

Local finiteness for real-virtual corrections to electroweak production in partonic collisions

We present a local subtraction scheme that enables the combined integration of loop momenta and the final-state parton phase space in real-virtual NNLO QCD corrections to cross sections for hadroproduction of electroweak and other colorless states. All initial- and final-state infrared singularities are subtracted at the integrand level in momentum space, yielding a locally finite integral ready for numerical integration in four dimensions. The subtraction terms are all based on the well-understood process of single-Higgs production. The core of our subtraction scheme relies on achieving local factorization in all infrared limits of real and virtual momenta. This necessitates systematic modifications of the original Feynman integrand for loop amplitudes, enabling gauge symmetry cancellations before performing integrations. Our approach provides an essential step toward NNLO cross-section calculations for hadron collider processes, where both loop and phase-space integrations are carried out numerically.

hep-ph↗

arXiv2026

Exploring Graph Learning Tasks with Pure LLMs: A Comprehensive Benchmark and Investigation

In recent years, large language models (LLMs) have emerged as promising candidates for graph tasks. Many studies leverage natural language to describe graphs and apply LLMs for reasoning, yet most focus narrowly on performance benchmarks without fully comparing LLMs to graph learning models or exploring their broader potential. In this work, we present a comprehensive study of LLMs on graph learning tasks, evaluating both off-the-shelf and instruction-tuned models across a variety of scenarios. Beyond accuracy, we discuss data leakage concerns and computational overhead, and assess their performance under few-shot/zero-shot settings, domain transfer, structural understanding, and robustness. Our findings show that LLMs, particularly those with instruction tuning, greatly outperform traditional graph learning models in few-shot settings, exhibit strong domain transferability, and demonstrate excellent generalization and robustness. Our study highlights the broader capabilities of LLMs in graph learning and provides a foundation for future research.

cs.LG↗

arXiv2026

Knowledge Homophily in Large Language Models

Large Language Models (LLMs) have been increasingly studied as neural knowledge bases for supporting knowledge-intensive applications such as question answering and fact checking. However, the structural organization of their knowledge remains unexplored. Inspired by cognitive neuroscience findings, such as semantic clustering and priming, where knowing one fact increases the likelihood of recalling related facts, we investigate an analogous knowledge homophily pattern in LLMs. To this end, we map LLM knowledge into a graph representation through knowledge checking at both the triplet and entity levels. After that, we analyze the knowledgeability relationship between an entity and its neighbors, discovering that LLMs tend to possess a similar level of knowledge about entities positioned closer in the graph. Motivated by this homophily principle, we propose a Graph Neural Network (GNN) regression model to estimate entity-level knowledgeability scores for triplets by leveraging their neighborhood scores. The predicted knowledgeability enables us to prioritize checking less well-known triplets, thereby maximizing knowledge coverage under the same labeling budget. This not only improves the efficiency of active labeling for fine-tuning to inject knowledge into LLMs but also enhances multi-hop path retrieval in reasoning-intensive question answering.

cs.LG↗

arXiv2025

Extending Graph Condensation to Multi-Label Datasets: A Benchmark Study

As graph data grows increasingly complicate, training graph neural networks (GNNs) on large-scale datasets presents significant challenges, including computational resource constraints, data redundancy, and transmission inefficiencies. While existing graph condensation techniques have shown promise in addressing these issues, they are predominantly designed for single-label datasets, where each node is associated with a single class label. However, many real-world applications, such as social network analysis and bioinformatics, involve multi-label graph datasets, where one node can have various related labels. To deal with this problem, we extends traditional graph condensation approaches to accommodate multi-label datasets by introducing modifications to synthetic dataset initialization and condensing optimization. Through experiments on eight real-world multi-label graph datasets, we prove the effectiveness of our method. In experiment, the GCond framework, combined with K-Center initialization and binary cross-entropy loss (BCELoss), achieves best performance in general. This benchmark for multi-label graph condensation not only enhances the scalability and efficiency of GNNs for multi-label graph data, but also offering substantial benefits for diverse real-world applications.

cs.LG↗

arXiv2025

Diagnosing and Addressing Pitfalls in KG-RAG Datasets: Toward More Reliable Benchmarking

Knowledge Graph Question Answering (KGQA) systems rely on high-quality benchmarks to evaluate complex multi-hop reasoning. However, despite their widespread use, popular datasets such as WebQSP and CWQ suffer from critical quality issues, including inaccurate or incomplete ground-truth annotations, poorly constructed questions that are ambiguous, trivial, or unanswerable, and outdated or inconsistent knowledge. Through a manual audit of 16 popular KGQA datasets, including WebQSP and CWQ, we find that the average factual correctness rate is only 57 %. To address these issues, we introduce KGQAGen, an LLM-in-the-loop framework that systematically resolves these pitfalls. KGQAGen combines structured knowledge grounding, LLM-guided generation, and symbolic verification to produce challenging and verifiable QA instances. Using KGQAGen, we construct KGQAGen-10k, a ten-thousand scale benchmark grounded in Wikidata, and evaluate a diverse set of KG-RAG models. Experimental results demonstrate that even state-of-the-art systems struggle on this benchmark, highlighting its ability to expose limitations of existing models. Our findings advocate for more rigorous benchmark construction and position KGQAGen as a scalable framework for advancing KGQA evaluation.

cs.CL↗

arXiv2025

Lessons Learned: A Multi-Agent Framework for Code LLMs to Learn and Improve

Recent studies show that LLMs possess different skills and specialize in different tasks. In fact, we observe that their varied performance occur in several levels of granularity. For example, in the code optimization task, code LLMs excel at different optimization categories and no one dominates others. This observation prompts the question of how one leverages multiple LLM agents to solve a coding problem without knowing their complementary strengths a priori. We argue that a team of agents can learn from each other's successes and failures so as to improve their own performance. Thus, a lesson is the knowledge produced by an agent and passed on to other agents in the collective solution process. We propose a lesson-based collaboration framework, design the lesson solicitation--banking--selection mechanism, and demonstrate that a team of small LLMs with lessons learned can outperform a much larger LLM and other multi-LLM collaboration methods.

cs.AI↗

arXiv2025

Composition of scalar mesons and their effects on nuclear matter properties in an extended linear sigma model

It has been argued that the iso-scalar and iso-vector mesons play significant roles in nuclear matter and neutron star structures. We improve the extended linear sigma model with baryons, proposed in our previous work, by introducing the flavor structures constructed from antisymmetric tensors of chiral representations to study these physics. The parameter space of this model is refined with well-reproduced nuclear matter properties at saturation density by the lowest order Lagrangian, ensuring consistency with vacuum results, such as $f_π\approx 134 \, \text{MeV}$. The anticipated plateau-like behaviors of the symmetry energy are predicted at intermediate densities, which is crucial for the consistency of GW170817 and the neutron skin thickness of $\text{Pb}^{208}$. Subsequently, neutron star structures are calculated using several parameter sets, and the results for the nuclear matter properties at saturation density align with empirical values. It is found that the neutron star structures are sensitive to the couplings between the iso-vector $a_0$ meson and nucleons and the four-vector meson couplings: small values of both are favorable. Meanwhile, nuclear matter properties at saturation density favor larger values of the latter and are not sensitive to the former. This signifies the statistical significance of neutron star observations when obtaining realistic chiral effective field theories or models at various densities. The parameter set favored by neutron star observations also aligns the behavior of the sound velocity with the conformal limit at high densities relevant to cores of massive stars. It is hoped that the results of this work can guide future studies on the relationship between the microscopic symmetry of strong interactions and macroscopic phenomena.

nucl-th↗