Searcharxiv⌕ Search

arXiv subjects

Rui Zou

Publications and source records attributed to Rui Zou.

At least 19 recordsLinked to original sources

AMTFV: Agentic Mathematical Tool-Flow Verification for LLM Self-Correction

Large language models have demonstrated strong mathematical problem-solving capabilities, yet reliably verifying their candidate answers remains challenging. Existing representative methods mainly revise outputs through natural-language reflection or assist verification by directly generating verification programs; the former may not reliably support exact computation, whereas the latter prematurely couples mathematical modeling with low-level implementation. We propose AMTFV (Agentic Mathematical Tool-Flow Verification). By introducing Mathematical Tool Flow (MTF) as an interrupt--execute--resume interface, AMTFV decouples verification modeling from concrete execution and supports exact computation through a mathematical toolbox. Specifically, the verification agent first constructs a verification workflow, encodes the mathematical objects and computational intent requiring reliable execution in an MTF request, and sends it to the mathematical toolbox agent. The latter parses the request, generates executable calls, and dispatches them to the backend for exact computation. Tool outputs then support candidate-answer adjudication, answer revision, and verification-workflow revision. We evaluate AMTFV on five challenging mathematical reasoning datasets with seven model configurations from DeepSeek, GPT, and Gemini. Experimental results show that AMTFV outperforms the representative baselines evaluated in this study overall; under an individual model configuration, it improves average accuracy over the strongest baseline by up to 8.3 percentage points, with larger gains on samples of medium and high verification complexity.

cs.AI↗

BiVerify: Closing the Loop Between Backward and Forward Verification for Mathematical Self-Correction

Large language models (LLMs) have made substantial progress in mathematical reasoning but still produce incorrect solutions to complex problems, indicating the need for more reliable self-correction mechanisms. Mathematical self-correction typically identifies errors through verification, which can proceed backward or forward. The two directions provide complementary verification capabilities. Backward verification checks a candidate answer against the original constraints, bypassing complex intermediate reasoning, but cannot ensure correct reasoning or objective completeness. Forward verification follows the problem conditions and current solution, enabling it to locate reasoning errors and assess objective completeness, but may overlook key constraints in complex solution chains. We propose BiVerify, a dynamically scheduled framework for bidirectional mathematical verification. BiVerify begins with backward verification to screen the answer against the original constraints, followed by forward verification of the reasoning process; subsequent direction switching, state resetting, and re-verification are governed by verification outcomes. Upon failure, the system revises the candidate or verification flow until the same candidate passes both directions. We also introduce shared mathematical tools that separate logical reasoning from precise computation. We systematically evaluate BiVerify on five challenging mathematical reasoning benchmarks using multiple mainstream DeepSeek, GPT, and Gemini models. BiVerify exceeds the strongest representative public baseline evaluated in this work by 13.5 percentage points in average accuracy, with the margin reaching 31.2 percentage points on GPT. Further analysis shows that BiVerify's overall performance exceeds that of either verification direction alone.

cs.AI↗

Agent-Guided Process Verification for Mathematical Self-Correction

Large language models have made substantial progress in mathematical reasoning, yet in complex multi-step reasoning, an error at any step may propagate along an incorrect path, and the resulting solution may omit necessary branches. Existing natural-language reflection methods often struggle to locate critical errors precisely, while outcome verification typically checks only whether the available candidate answers satisfy the problem conditions and can therefore overlook necessary branches. To address these limitations, we propose Agent-Guided Process Verification (AGPV) for mathematical self-correction. AGPV decomposes the current solution into an ordered sequence of key claims, verifies them individually while invoking a shared mathematical toolbox when needed, and stops upon detecting the first error. A summarization agent converts the failure evidence into a diagnosis that guides a correction agent to re-derive the subsequent solution from that point; AGPV then adapts the next verification iteration to the revised solution. AGPV also verifies solution completeness. We further prove that AGPV guarantees solution correctness and completeness when key claims provide sufficient coverage and local verification is reliable, provided that all key claims pass verification. The implementation compresses long working contexts and reuses successful tool results within each problem. We evaluate AGPV on five challenging mathematical reasoning datasets with four models from DeepSeek, Qwen, and GPT. AGPV achieves an average accuracy of 63.8%, outperforming the strongest among the representative baselines evaluated (50.3%) by 13.5 percentage points, with higher repair coverage than the programmatic-verification baselines.

cs.AI↗

Hawk: Harnessing Hardware-Aware Knowledge for High-Performance NPU Kernel Generation

Developing high-performance kernels for Neural Processing Units (NPUs) is a critical industry bottleneck, requiring developers to manually navigate implicit hardware constraints and strict memory hierarchies. While large language models offer immense automation potential, they fail catastrophically on NPUs due to a fundamental lack of hardware-specific priors. Naively transplanting code snippets from similar NPU kernels may pass the compiler, but it consistently triggers runtime crashes and performance degradation by blindly violating underlying hardware constraints. To overcome this, we introduce Hawk, a training-free framework that harnesses hardware-aware knowledge through three core modules: (1) Run-Time Knowledge Synthesis Module, which employs a Triple-Part Executable Knowledge Representation to inherently couple the error context with executable semantics; (2) Bottleneck-Aware Knowledge Retrieval Module, which implements a 2D-Retrieval paradigm to project queries into orthogonal syntactic and hardware-aligned semantic spaces; and (3) Effect-Driven Knowledge Distillation Module, which leverages LLM-driven semantic arbitration to continuously distill the knowledge by pruning errors and consolidating redundancies based on the empirical execution feedback. Extensive evaluations on real-world NPU workloads demonstrate that Hawk elevates generation accuracy from 49.4% to 80.0%, while achieving up to a 2.2x execution speedup over state-of-the-art baselines.

cs.AI↗

Inertia Matching Principle: Improving Transient Synchronization Stability in Hybrid Power Systems With VSGs and SGs

This paper investigates the transient synchronization stability in power systems hybridized with virtual synchronous generators (VSGs) and synchronous generators (SGs). A relative swing equation model is established to capture the transient synchronization dynamics between the VSG and the SG. Based on this model, both static and dynamic characteristics are systematically analyzed, and a quantitative stability level index is derived to elucidate the underlying stability mechanism. Then, two fundamental inertia matching principles are identified. First, a new instability mechanism induced by improper inertia matching between the VSG and the SG is revealed. It is identified that increasing the VSG's inertia does not monotonically improve transient stability, as commonly presumed. Instead, an optimal inertia matching constant exists that maximizes stability performance. Second, the influence of the VSG share on the synchronization stability is discovered to be strongly influenced by the matching between the VSG's inertia level and its voltage strength (i.e., output impedance). To achieve reliable and robust synchronization stability, proper coordination between the VSG's inertia and virtual impedance is essential. Finally, a coordinated stabilization strategy based on inertia matching and virtual impedance adjustment is proposed to enhance transient synchronization stability performance while suppressing fault current. Simulations conducted on a two-machine system and the IEEE 39-bus system validate the theoretical findings and demonstrate the effectiveness of the proposed strategy.

eess.SY↗

Stable manifolds, Horseshoes and Lyapunov exponents for $C^1$ diffeomorphisms without domination

We develop the nonuniformly hyperbolic theory for $C^1$ diffeomorphisms admitting continuous invariant splitting without domination. This framework includes stable manifold theorems, shadowing and closing lemmas, the existence of horseshoes and the approximation of Lyapunov exponents. The foundation is a new family of resonance blocks, each arising as the forward limit set of a typical point at carefully chosen resonance times where expansion, contraction and a weak scale-dependent domination coexist.

math.DS↗

Self-Error Adjustment: Theory and Practice of Balancing Individual Performance and Diversity in Ensemble Learning

Ensemble learning boosts performance by aggregating predictions from multiple base learners. A core challenge is balancing individual learner accuracy with diversity. Traditional methods like Bagging and Boosting promote diversity through randomness but lack precise control over the accuracy-diversity trade-off. Negative Correlation Learning (NCL) introduces a penalty to manage this trade-off but suffers from loose theoretical bounds and limited adjustment range. To overcome these limitations, we propose a novel framework called Self-Error Adjustment (SEA), which decomposes ensemble errors into two distinct components: individual performance terms, representing the self-error of each base learner, and diversity terms, reflecting interactions among learners. This decomposition allows us to introduce an adjustable parameter into the loss function, offering precise control over the contribution of each component, thus enabling finer regulation of ensemble performance. Compared to NCL and its variants, SEA provides a broader range of effective adjustments and more consistent changes in diversity. Furthermore, we establish tighter theoretical bounds for adjustable ensemble methods and validate them through empirical experiments. Experimental results on several public regression and classification datasets demonstrate that SEA consistently outperforms baseline methods across all tasks. Ablation studies confirm that SEA offers more flexible adjustment capabilities and superior performance in fine-tuning strategies.

cs.LG↗

Hide and Seek with LLMs: An Adversarial Game for Sneaky Error Generation and Self-Improving Diagnosis

Large Language Models (LLMs) excel in reasoning and generation across domains, but still struggle with identifying and diagnosing complex errors. This stems mainly from training objectives that prioritize correct answers, limiting exposure to and learning from errors. While recent studies have begun to address this by introducing error signals, most rely on shallow, static errors, restricting improvement in deep diagnostic ability. To overcome this, we propose Hide and Seek Game (HSG), a dynamic adversarial framework for error generation and diagnosis, and evaluate it on mathematical problem-solving. HSG involves two adversarial roles: Sneaky, which "hides" by generating subtle, deceptive reasoning errors, and Diagnosis, which "seeks" to accurately detect them. Through adversarial co-evolution, both error stealth and diagnostic precision are enhanced. Experiments on several math reasoning tasks show that HSG significantly boosts error diagnosis, achieving 16.8\%--31.4\% higher accuracy than baselines like GPT-4o. We also release a challenging dataset of deceptive errors and diagnostic annotations as a benchmark for future research.

cs.AI↗

A variational principle for nonlinear local pressure

In this paper, we introduce a concept of nonlinear local topological pressure defined via open covers and establish a corresponding variational principle. Furthermore, we provide multiple equivalent characterizations of nonlinear pressure using different cover-based approaches.

math.DS↗

Design, Construction, and Testing of the APOLLO ATCA Blades for Use at the HL-LHC

The Apollo Advanced Telecommunications Computing Architecture (ATCA) platform is an open-source design consisting of a generic "Service Module" (SM) and a customizable "Command Module" (CM), allowing for cost-effective use in applications such as the readout of the inner tracker and the Level-1 track trigger for the CMS Phase-II upgrade at the HL-LHC. The SM integrates an intelligent IPMC, robust power entry and conditioning systems, a powerful system-on-module computer, and flexible clock and communication infrastructure. The CM is designed around two Xilinx Ultrascale+ FPGAs and high-density, high-bandwidth optical transceivers capable of 25 Gb/s. Crates of Apollo blades are currently being tested at Boston University, Cornell University, and CERN.

hep-ex↗

Gradual Vigilance and Interval Communication: Enhancing Value Alignment in Multi-Agent Debates

In recent years, large language models have shown exceptional performance in fulfilling diverse human needs. However, their training data can introduce harmful content, underscoring the necessity for robust value alignment. Mainstream methods, which depend on feedback learning and supervised training, are resource-intensive and may constrain the full potential of the models. Multi-Agent Debate (MAD) offers a more efficient and innovative solution by enabling the generation of reliable answers through agent interactions. To apply MAD to value alignment, we examine the relationship between the helpfulness and harmlessness of debate outcomes and individual responses, and propose a MAD based framework Gradual Vigilance and Interval Communication (GVIC). GVIC allows agents to assess risks with varying levels of vigilance and to exchange diverse information through interval communication. We theoretically prove that GVIC optimizes debate efficiency while reducing communication overhead. Experimental results demonstrate that GVIC consistently outperforms baseline methods across various tasks and datasets, particularly excelling in harmfulness mitigation and fraud prevention. Additionally, GVIC exhibits strong adaptability across different base model sizes, including both unaligned and aligned models, and across various task types.

cs.AI↗

Revolutionizing MRI Data Processing Using FSL: Preliminary Findings with the Fugaku Supercomputer

The amount of Magnetic resonance imaging data has grown tremendously recently, creating an urgent need to accelerate data processing, which requires substantial computational resources and time. In this preliminary study, we applied FMRIB Software Library commands on T1-weighted and diffusion-weighted images of a single young adult using the Fugaku supercomputer. The tensor-based measurements and subcortical structure segmentations performed on Fugaku supercomputer were highly consistent with those from conventional systems, demonstrating its reliability and significantly reduced processing time.

physics.med-ph↗

Time delay of fast radio burst population with respect to the star formation history

In spite of significant progress in the research of fast radio bursts (FRBs) in recent decade, their origin is still under extensive debate. Investigation on the population of FRBs can provide new insight into this interesting problem. In this paper, based on the first CHIME/FRB catalog, we construct a Bayesian framework to analyze the FRB population, with the selection effect of the CHIME telescope being properly taken into account. The energy function is modeled as the power-law with an exponential cutoff. Four redshift distribution models are considered, i.e., the star formation history (SFH) model, and three time-delayed models (Gaussian delay, log-normal delay, and power-law delay). The free parameters are simultaneously constrained using Bayesian inference method, and the Bayesian information criterion (BIC) is used in model comparison. According to BIC, the log-normal delay model fits the data best. The power-law delay model and Gaussian delay model can also give reasonable fits, although they are not as good as the log-normal delay model. However, the SFH model is strongly disfavored compared with the three time-delayed models. The energy function is tightly constrained and is almost independent of the redshift models, with the best-fitting power-law index $α\approx 1.8$, and cut-off energy $\log(E_c/{\rm erg})\approx 42$. The FRB population shows on average $3\sim 5$ billion years time delay with respect to the SFH. Therefore, the hypothesis that the FRB population traces the SFH is conclusively ruled out.

astro-ph.HE↗

Revised Constraints on the fast radio burst population from the first CHIME/FRB catalog

In this paper, we investigate the FRB population using the first CHIME/FRB catalog. We first reconstruct the extragalactic dispersion measure -- redshift relation ($\mathrm{DM_E} - z$ relation) from well-localized FRBs, then use it to infer redshift and isotropic energy of the first CHIME/FRB catalog. The intrinsic energy distribution is modeled by the power law with an exponential cutoff, and the selection effect of the CHIME telescope is modeled by a two-parametric function of specific fluence. For the intrinsic redshift distribution, the star formation history (SFH) model, as well as other five SFH-related models are considered. We construct the joint likelihood of fluence, energy and redshift, and all the free parameters are constrained simultaneously using Bayesian inference method. The Bayesian information criterion (BIC) is used to choose the model that best matches the observational data. For comparison, we fit our models with two data samples, i.e. the Full sample and the Gold sample. The power-law index and cutoff energy are tightly constrained to be $1.8 \lesssim α\lesssim 1.9$ and $\mathrm{log}(E_c/{\rm erg}) \approx 42$, which are almost independent of the redshift distribution model and the data sample we choose. The parameters involving the selection effect strongly depends on the data sample, but are insensitive to the redshift distribution model. According to BIC, the pure SFH model is strongly disfavored by both the Full sample and Gold sample. For the rest five SFH-related redshift distribution models, most of them can match the data well if the parameters are properly chosen. Therefore, with the present data, it is still premature to draw a conclusive conclusion on the FRB population.

astro-ph.HE↗

Multiplayer Homicidal Chauffeur Reach-Avoid Games: A Pursuit Enclosure Function Approach

This paper presents a multiplayer Homicidal Chauffeur reach-avoid differential game, which involves Dubins-car pursuers and simple-motion evaders. The goal of the pursuers is to cooperatively protect a planar convex region from the evaders, who strive to reach the region. We propose a cooperative strategy for the pursuers based on subgames for multiple pursuers against one evader and optimal task allocation. We introduce pursuit enclosure functions (PEFs) and propose a new enclosure region pursuit (ERP) winning approach that supports forward analysis for the strategy synthesis in the subgames. We show that if a pursuit coalition is able to defend the region against an evader under the ERP winning, then no more than two pursuers in the coalition are necessarily needed. We also propose a steer-to-ERP approach to certify the ERP winning and synthesize the ERP winning strategy. To implement the strategy, we introduce a positional PEF and provide the necessary parameters, states, and strategies that ensure the ERP winning for both one pursuer and two pursuers against one evader. Additionally, we formulate a binary integer program using the subgame outcomes to maximize the captured evaders in the ERP winning for the pursuit task allocation. Finally, we propose a multiplayer receding-horizon strategy where the ERP winnings are checked in each horizon, the task is allocated, and the strategies of the pursuers are determined. Numerical examples are provided to illustrate the results.

eess.SY↗

YAYI 2: Multilingual Open-Source Large Language Models

As the latest advancements in natural language processing, large language models (LLMs) have achieved human-level language understanding and generation abilities in many real-world tasks, and even have been regarded as a potential path to the artificial general intelligence. To better facilitate research on LLMs, many open-source LLMs, such as Llama 2 and Falcon, have recently been proposed and gained comparable performances to proprietary models. However, these models are primarily designed for English scenarios and exhibit poor performances in Chinese contexts. In this technical report, we propose YAYI 2, including both base and chat models, with 30 billion parameters. YAYI 2 is pre-trained from scratch on a multilingual corpus which contains 2.65 trillion tokens filtered by our pre-training data processing pipeline. The base model is aligned with human values through supervised fine-tuning with millions of instructions and reinforcement learning from human feedback. Extensive experiments on multiple benchmarks, such as MMLU and CMMLU, consistently demonstrate that the proposed YAYI 2 outperforms other similar sized open-source models.

cs.CL↗

A Measurement of Hubble Constant Using Cosmographic Approach from Fast Radio Bursts and SNe Ia

The Hubble constant ${H}_0$ is a crucial parameter in cosmology. However, different cosmic observations have resulted in varying posterior results for ${H}_0$, leading to what is known as the ${H}_0$ tension. In order to address this issue, it is beneficial to use other dataset to constrain ${H}_0$. In this paper, via the cosmographic approach based on the Friedman-Lemaitre-Robertson-Walker (FLRW) metric to the dispersion measure of the intergalactic medium ${\rm{DM}}_{\rm{IGM}}(z)$ of Fast Radio Bursts (FRBs), we obtain the Taylor expansion of $\langle{\rm{DM}}_{\rm{IGM}}(z)\rangle$ in terms redshift $z$. The result for Hubble constant $H_0=65.5^{+6.4}_{-5.4}$ ${\rm{km~s^{-1}~Mpc^{-1}}}$ $(68$$\%$ ${\rm{C.L.}}) $, cosmological deceleration parameter $q_0=-0.50\pm 0.20 $ and the jerk parameter $j_0=-0.1^{+2.0}_{-2.5}$ using uncalibrated Supernova Ia (SNe Ia) Pantheon dataset combined with 18 localized FRBs are obtained. To demonstrate the impact of parameter degeneracies on our analysis methods, we compare the results using three different forms of $f_{\rm{IGM}}(z)$ and two different prior distributions for $Ω_{\rm{b,0}}$. Then we find that the uncertainty in $H_0$ is not significantly affected by the prior range of $f_{\rm{IGM}}(z)$ and $Ω_{\rm{b,0}}$, but the mean value is influenced by the priors for $f_{\rm{IGM}}(z)$ and $Ω_{\rm{b,0}}$ due to parameter degeneracies with $H_0$. Employing $f_{\rm{IGM}}(z)$ that evolves with redshift, we obtain the constraints for $H_0=69.0^{+6.7}_{-5.7}$ ${\rm{km~s^{-1}~Mpc^{-1}}}$. Furthermore, the mock analyses give a posterior estimation of $H_0$ with an accuracy of 4.6\% and higher precision for $q_0$ and $j_0$ in the near future.

astro-ph.CO↗