arXiv subjects

Xin Peng

Publications and source records attributed to Xin Peng.

At least 19 recordsLinked to original sources

arXiv2026

MeClear: Cooperative Game-Theoretic Attribution and Risk-Aware Memory Clearance for Long-Horizon LLM Agents

Long horizon Large Language Model (LLM) agents rely on external memory systems to preserve user preferences and task knowledge across extended interactions. Conventional retrieval mechanisms optimize semantic compatibility rather than downstream utility, frequently introducing outdated, misleading, or conflicting evidence into the active context. We present MeClear, a task conditioned memory clearance framework that identifies memories featuring negative downstream utility through cooperative attribution and selectively suppresses them from agent execution. MeClear combines Leave One Out screening with sampled cooperative Shapley attribution to distribute utility across interacting evidence, effectively resolving redundant conflict masking where single removal evaluations fail. Utilizing attribution rankings, MeClear executes a query scoped minimal clearance strategy over a nested filtration, verifying task recovery on the cleared context without permanently altering the persistent memory bank. Comprehensive experimental evaluations across ten long dialogue memory pools demonstrate that MeClear achieves a target recall of 85.9% and an overall task recovery rate of 82.3%, representing a 25.5 percentage point improvement over Leave One Out (LOO) baselines.

cs.AI↗

arXiv2026

Trace2Tower: Transition-Aware EigenTrace Induction of Multi-Level Skills for LLM Agents

Large language model agents increasingly rely on execution traces to master complex interactive tasks. However, current paradigms are bottlenecked by shallow trajectory retrieval and flat skill summarization, fundamentally ignoring the temporal dependencies and outcome-conditioned topology of agent behavior. We introduce Trace2Tower, a transition-aware EigenTrace framework that distills raw trajectories into a robust skill hierarchy. Trace2Tower abstracts step-level interactions into canonical events, constructing a unified graph governed by semantic compatibility, transition dynamics, and outcome evidence. Through a novel contrastive spectral decomposition, it isolates stable, success-aligned behavioral modes while rigorously suppressing failure-prone shortcuts. These modes organically populate a dynamic skill tower of action templates, procedural routines, and overarching task strategies, continuously refined via verifier-guided feedback. On ALFWorld, Trace2Tower achieves 87.31% success requiring only 10.35 steps and 0.26 invalid actions; on WebShop, it reaches 50.67% exact success. Across both benchmarks, Trace2Tower significantly outperforms existing baselines in task mastery and context-efficient experience reuse.

cs.AI↗

arXiv2026

OpenHarmony Bench: Evaluating LLMs and Coding Agents on OpenHarmony App Development

We present OPENHARMONY BENCH, an app-level coding benchmark for evaluating LLM-based coding agents on OpenHarmony ArkTS applications. Unlike function-level benchmarks, it evaluates complete app-level changes: each task requires an agent to modify a buildable ArkTS project so that a requested behavior works end to end, involving UI state, data persistence, build configuration, and platform APIs. The benchmark installs and drives the delivered application on a device to check whether the behavior is observable. It covers three input sources: natural-language feature requests (new-feature), structured scenario specifications (spec-driven), and bug descriptions (bug-fix). The benchmark contains 153 top-level tasks and 242 Feature points (F-points), where an F-point is one executable behavior check. The snapshot includes 32 new-feature tasks, 50 spec-driven tasks with 139 F-points, and 71 bug-fix tasks. The main leaderboard is scored over top-level tasks rather than independently weighted F-points. We describe the benchmark construction, statistics, and build-and-test evaluation pipeline, and evaluate DevEco Code with eight LLMs across three independent full-suite runs per configuration. Three findings emerge. First, newer generations complete more tasks than their predecessors within evaluated model-family pairs. Second, buildability is close to saturated while behavioral correctness is not: mean Final Build Success Rate is 94.77% to 100.00%, whereas mean Task Completion is 48.36% to 58.39%. Third, spec-driven tasks have the lowest Task Completion under all-checks task scoring, with no configuration exceeding 35%. The code, data, tasks, reference solutions, tests, evaluation scripts, and leaderboard are released through the official OPENHARMONY BENCH website at https://bench.matrix.openharmony.cn/.

cs.SE↗

arXiv2026

SkillSentry: Reliable Skill Execution for LLM Agents via Runtime Assurance

LLM agents are increasingly equipped with skills to perform complex tasks through multi-step reasoning and tool use. Although skills provide reusable procedural knowledge, agents may still execute them unreliably. Even when an agent has demonstrated the capability to complete tasks under the guidance of a skill, it may fail to do so consistently across similar tasks or repeated runs due to deviations from the skill procedure or incorrect execution of individual steps. Such instability limits the practical reliability of LLM agents. To address this problem, we propose SkillSentry, a skill-oriented runtime assurance framework built upon a new domain-specific language (DSL) for representing runtime guidance for skill execution. SkillSentry initializes the runtime guidance by combining a skill specification extracted from the corresponding skill document with execution experience mined from historical successful and failed traces. It then wraps around the agent execution loop to monitor and guide skill execution under the current guidance, while iteratively refining the guidance using newly collected traces. We evaluate SkillSentry on 15 skills across two LLM agents, each paired with two backbone models, i.e., Claude Code with Claude-Haiku-4.5 and Claude-Opus-4.6, and Codex with GPT-5.2 and GPT-5.4. Our results show that SkillSentry improves the task success rate of LLM agents by 24.1% across skills, on average, while exhibiting lower variability across repeated runs.

cs.AI↗

arXiv2026

DOCSCHISEL: Adaptive Tool Documentation Optimization Framework for LLM Agents

Large language models (LLMs) increasingly rely on external tools to accomplish complex real-world tasks, making tool documentation a critical grounding resource for LLM agents. Existing studies mainly focus on improving the tool-use capabilities of LLM agents, while largely treating tool documentation as a fixed input. Although several recent works attempt to optimize tool documentation through rewriting or compression, little is known about how the information contained in tool documentation affects agent performance across different settings. To bridge this gap, we conduct a large-scale empirical study on tool documentation for LLM agents. Our study reveals substantial heterogeneity in the information fields provided by existing tool documentation. Moreover, the effectiveness of different information fields is highly dependent on the task domain, LLM backbone, and agent paradigm, indicating that no fixed tool documentation can consistently generalize across diverse agent settings. Motivated by these findings, we propose DocsChisel, an adaptive tool documentation optimization framework for LLM agents. DocsChisel analyzes failed execution traces of a target LLM agent to identify documentation-related issues, and iteratively optimizes tool documentation by adding, removing, and refining information fields for each tool. We evaluate DocsChisel against two state-of-the-art baselines, i.e., EasyTool and DRAFT. Experimental results show that DocsChisel improves the task success rate of LLM agents by 95.89% over the original tool documentation and by 75.15%, on average, over existing baselines, while incurring limited optimization time and token overhead

cs.LG↗

arXiv2026

FlowScout: From Execution Feedback to Reliable Tool-Using Agent Workflows

Agentic workflows have become an important abstraction for building reliable LLM-based automation systems by organizing large language models (LLMs), tools, and control logic into explicit execution structures. However, constructing high-quality agentic workflows remains largely manual and requires substantial domain expertise. Recent studies have explored automatic agentic workflow generation from historical task-solving records, but they mainly produce LLM-centric workflows, where real tool executions are abstracted and simulated by LLM nodes, limiting the usability and stability of generated workflows. To address these limitations, we propose FlowScout, an execution-guided framework for generating tool-integrated agentic workflows from historical task-solving records. Specifically, FlowScout represents an agentic workflow as a directed graph composed of LLM nodes, tool-calling nodes, and dependency edges. It first mines a common tool coordination skeleton from historical records to construct an initial workflow, and then refines the workflow topology through Monte Carlo tree search guided by execution feedback. We evaluate FlowScout on four representative task domains and compare it with three baselines, i.e., PM4Py, ReAct and AFlow. Experimental results show that agentic workflows generated by FlowScout improve tool invocation correctness by at least 92.69% and execution quality by at least 17.66% over the baselines, while achieving lower performance variation across repeated runs.

cs.LG↗

arXiv2026

RESTOR: Automated Test Oracle Generation for RESTful APIs via Reinforcement Learning

Modern REST API testing faces a critical challenge in defining reliable test oracles, particularly in agile industrial environments where formal specifications (e.g., OpenAPI) are frequently missing or outdated, and historical execution logs are unavailable for newly deployed endpoints. In this paper, we present Restor (Reinforcement Enhanced Single-Traffic Oracle generator for REST APIs), a framework that generates executable test assertions from a single observed request-response pair in a black-box setting. Unlike existing approaches that rely on rule-based templates or massive training logs, Restor utilizes a novel data augmentation pipeline to fine-tune a lightweight Large Language Model (LLM) via Group Relative Policy Optimization (GRPO). This training process enables the model to internalize testing "common sense" by optimizing a reward function that jointly encourages: (i) the selection of stable, semantically meaningful fields for validation and the avoidance of dynamic noise (e.g., timestamps or trace IDs); (ii) the generation of robust assertions that withstand logic variations. We evaluate Restor on an industrial dataset comprising over 2,300 API traces across 246 real-world services. Comprehensive experiments demonstrate that Restor significantly outperforms prompt-engineered baselines and generalist models, achieving a superior $F_1$ score of 85.42% in key field identification and increasing the proportion of semantically accurate assertions. Furthermore, deployment in a production CI/CD workflow at ByteDance confirms its practical value: the system raised the adoption rate of automatically generated test cases from 74.1% to over 96%, substantially reducing manual Quality Assurance (QA) effort while ensuring high execution stability.

cs.SE↗

arXiv2026

Industrial Practice of LLM-Based Test Case Carving and Assertion Generation (Experience Paper)

Enterprise regression testing for microservice systems is often constrained by incomplete or outdated documentation. In practice, QA engineers frequently rely on real execution traffic to reconstruct business scenarios; however, turning raw traffic into replayable regression tests with stable validation logic remains labor-intensive and error-prone. This paper presents NL2Test, an end-to-end approach and tool that generates executable API regression tests from (i) a natural-language scenario description and (ii) a traffic capture recorded while executing the scenario. NL2Test addresses two coupled tasks: test case carving, which extracts a minimal replayable request sequence and reconstructs data dependencies so that dynamic values are bound from their responses rather than hard-coded; and assertion generation, which produces assertions aligned with business intent while avoiding non-deterministic fields and hallucinated paths. To improve reliability, NL2Test uses LLMs for semantic interpretation and constrained code synthesis, and uses deterministic algorithms for request filtering, dependency confirmation via value consistency, and assertion-path validation. We evaluate NL2Test on 51 industrial regression scenarios extracted from a large consumer-facing Internet company. NL2Test achieves an exact-match rate of 82.4% (42/51), and produces a functionally usable draft in 98.0% (50/51) of scenarios when allowing minor post-edits. In a 9-month production deployment starting in March 2025, NL2Test generated 3,196 test cases with an overall code adoption rate of 85.4%. These results indicate that traffic-grounded generation with deterministic guardrails can substantially reduce manual effort while improving regression automation in complex microservice environments.

cs.SE↗

arXiv2026

ProfMalPlus: Agent-Coordinated Detection of Malicious NPM Packages via Static-Dynamic Analysis Synergy

Open source software is vulnerable to supply-chain attacks through transitive dependencies, especially malicious code injected into NPM packages. Existing detectors often inadequately model obfuscated behavior, overlook JavaScript's object-centric features, poorly coordinate static and dynamic analysis, and lose semantic information during behavior abstraction. We propose ProfMalPlus, a malicious NPM package detector combining object-sensitive behavior graphs with coordinated LLM reasoning over annotated code slices. It identifies installation commands and entry files, then constructs graphs capturing sensitive APIs, third-party calls, and unresolved calls. From these graphs, ProfMalPlus extracts security-relevant slices and adds inline static analysis evidence. Local judge agents independently assess each slice. Self-consistency consolidates repeated judgements to reduce LLM variance, while a global judge synthesizes their reports into an entry-level verdict. For undetermined cases, a router selects either third-party enrichment, which adds registry derived module and method semantics, or dynamic augmentation, which executes the package in a sandbox to resolve runtime dependent behavior. The enriched evidence is fed back for reassessment. Finally, a localization agent reports malicious code snippets with explanations. ProfMalPlus achieves a 98.1% F1-score, outperforming state-of-the-art detectors by 3.5% to 52.6%. It also identified 597 previously unknown malicious packages, all confirmed and removed from NPM.

cs.SE↗

arXiv2026

x-Prediction Is All You Need:Training-Free Accelerated Generation via Endpoint Decodability

Diffusion and flow matching models generate high-quality samples, but their ODE samplers often need tens to hundreds of neural function evaluations (NFEs). This remains a practical challenge for released checkpoints, since many accelerators require additional design choices and training cost through retraining, distillation, or trajectory redesign. We investigate a different route based on $x$-prediction. During sampling, standard affine probability paths already expose $x_0$ information: an intermediate state and its path velocity determine a principled estimate of the clean sample. We formalize this property as \textbf{endpoint decodability} and show that the decoder is the minimum-MSE estimator $\mathbb{E}[x_0\mid x_t]$ under the usual $\ell_2$ objective. This yields \textbf{Truncated Jump Sampling} (TJS): stop the ODE at an early-exit time $t^*$ and return the decoded $x_0$. TJS requires no retraining, distillation, or architecture change. Across SDXL, SD3.5M, Z-Image-Turbo, and three class-conditional benchmarks, it reduces NFEs by 20--70\% with near-matched quality. The analysis also shows why endpoint prediction can work without straightening the trajectory, providing inference acceleration without trajectory redesign.

cs.LG↗

arXiv2026

LogCopilot: Automating Log Aggregation Analysis through Large Language Models

Logs record the runtime behavior of software and are widely used in various tasks such as debugging, testing, and fault diagnosis. With the increase in system size and complexity, log analysis has gradually become a challenging task. Current industrial systems typically use log aggregation systems such as Grafana Loki and ELK to simplify the log collection and analysis process. Engineers write queries using the DSL query language provided by these systems can complete a variety of log analysis tasks. However, writing these queries is often time-consuming and labor-intensive, as it requires engineers to have a thorough understanding of the DSL syntax and the detailed information contained in the logs. To address these challenges, this paper proposes LogCopilot, an automated log aggregation analysis framework based on large language models (LLMs). LogCopilot accepts natural language log analysis instructions and accomplishes automated log analysis through knowledge retrieval and tool calling. LogCopilot constructs a hierarchical knowledge base to represent and provide key knowledge in logs. And it achieves automated log aggregation analysis by generating and executing LogQL queries. The evaluation based on four log datasets confirm the effectiveness of LogCopilot, which achieves an average accuracy of 76.8% and outperforms baseline approaches. Moreover, experiment results shows that LogCopilot is effective in LogQL query generation.

cs.SE↗

arXiv2026

Prior-Averaged Ranking of Low-Order Monomial Potentials in Low-Temperature Warm Inflation

We compare the relative prior-averaged weights of the monomial potentials $V_p(\phi)=\lambda_p\phi^p/p$, with $p=2,3,4$, in low-temperature warm inflation with the dissipative coefficient fixed to $\Upsilon=C_\phi T^3/\phi^2$. The dissipative form, background equations, primordial-spectrum prescription, and compressed observable likelihood are held fixed while the monomial power is varied. For each branch, the warm background equations, including radiation backreaction, are solved numerically, and the broadened likelihood for $(A_s,n_s,r_{0.05})$ is integrated over the adopted prior domain to obtain $Z_{\rm eff}^{(A_s,n_s,r)}$. For the reference configuration $N_*=55$, $\sigma_r=0.005$, and the viability-conditioned prior A, we obtain $\Delta\ln Z_{\rm eff}(p=2)=-32.18$ and $\Delta\ln Z_{\rm eff}(p=3)=-6.99$ relative to $p=4$. The same ranking is retained for the specific variations of $N_*$, the prior domains, the random seeds, and the smoothing width of the upper-bound penalty examined here. A representative quartic trajectory gives $n_s=0.96420$, $r_{0.05}=0.02663$, $Q_*=4.68\times10^{-3}$, and $T_*/H_*=10.67$. Its scalar spectrum is evaluated using the assumed Bose--Einstein occupation prescription. A component test at this representative point indicates that the occupation term contributes more to the scalar enhancement than the fitted dissipative growth factor. Thus, within the compressed-likelihood setup, the adopted thermal-occupation prescription, and the adopted viability-conditioned priors, the prior-averaged ranking is $p=4>p=3\gg p=2$.

astro-ph.CO↗

arXiv2026

Three Heads Are Better Than One: A Multi-perspective Reasoning Framework for Enhanced Vulnerability Detection

Automated vulnerability detection is crucial for enhancing software security by identifying potential flaws that attackers could exploit, thereby reducing the reliance on labor-intensive manual code audits. Recent advancements have shifted towards leveraging large language models (LLMs) for vulnerability detection, with techniques like Vul-RAG and VulnSage demonstrating progress through structured prompting and external knowledge integration. However, these approaches typically rely on a single reasoning paradigm, limiting their ability to address the complex and diverse nature of real-world vulnerabilities. To overcome these limitations, we propose ReasonVul, a novel multi-perspective reasoning framework that harnesses cognitive synergy among three specialized LLM agents, each embodying a distinct reasoning mode. The framework begins with independent analyses of the source code, followed by a structured debate mechanism to resolve conflicts through iterative rebuttal and revision, ultimately converging on a collaborative judgment. Evaluated on the PrimeVul dataset, ReasonVul achieves a PairAcc of 40.00% and an F1-score of 72.52%, surpassing the best baseline by 81.24% in PairAcc. Further tests on the JITVUL dataset confirm its generalizability, with a PairAcc of 28.67%. Additionally, we analyzed 542 conflict cases and found that 389 were correctly resolved, highlighting the framework's ability to uncover hidden vulnerabilities through the error-correction mechanism driven by the debate. This work emphasizes the importance of multi-perspective reasoning and collaborative validation in achieving robust and comprehensive vulnerability detection in real-world software systems.

cs.SE↗

arXiv2026

Structure of Torus Fibration Under the First Betti Number Restriction

We study torus bundles with affine structure groups. First, we establish a rigidity result under constraints on the first Betti numbers: If $ \text{b}_{1}(M)-\text{b}_{1}(N)=\dim M-\dim N $ holds for a torus bundle $M$ with an affine structure group over a closed manifold $N$, then $M$ can be classified. Second, we obtain some necessary and sufficient conditions for the topological splitting of principal torus bundles. These results improve the understanding of the geometry of collapsing sequences under the first Betti number constraints, thereby extending the prior work by Huang-Wang.

math.DG↗

arXiv2026

P-Guide: Parameter-Efficient Prior Steering for Single-Pass CFG Inference

Classifier-Free Guidance (CFG) is essential for high-fidelity conditional generation in flow matching, yet it imposes significant computational overhead by requiring dual forward passes at each sampling step. In this work, we address this bottleneck by introducing \textbf{P-Guide}, a framework that achieves high-quality guidance through a single inference pass by modulating only the initial latent state. We further show that, under a first-order approximation, P-Guide is equivalent to CFG in the sense that it steers generation from the prior space, without requiring explicit velocity field extrapolation during sampling. We consider both homoscedastic and \textbf{heteroscedastic} priors, and find that jointly modeling the mean and variance enables adaptive loss attenuation and improved robustness to data uncertainty. Extensive experiments demonstrate that P-Guide reduces inference latency by approximately 50\% while maintaining fidelity and prompt alignment competitive with standard dual-pass CFG baselines.

cs.AI↗

arXiv2026

Flow Matching with Arbitrary Auxiliary Paths

We introduce a new generative modeling framework, \textbf{Flow Matching with Arbitrary Auxiliary Paths (AuxPath-FM)}, which generalizes conditional flow matching by incorporating an auxiliary variable drawn from an arbitrary distribution into the probability path. Unlike prior methods that restrict auxiliary components to Gaussian noise, AuxPath-FM allows the variable $\eta$ to follow any distribution, producing trajectories of the form $X_t = a(t)X_1 + b(t)X_0 + c(t)\eta$. We theoretically demonstrate that this construction preserves the continuity equation and maintains a training objective consistent with the marginal formulation. This flexibility enables the design of diverse probability paths using various priors, including Gaussian, Uniform, Laplace, and discrete Rademacher distributions, each offering unique geometric properties for generative flows. Furthermore, our framework allows for specialized tasks such as label-guided generation by encoding structured semantic information into the auxiliary distribution. Overall, AuxPath-FM provides a principled and general foundation for probability path design, offering both theoretical generality and practical flexibility for diverse generative modeling tasks.

cs.LG↗

arXiv2026

Bridging Generation and Training: A Systematic Review of Quality Issues in LLMs for Code

Large language models (LLMs) frequently generate defective outputs in code generation tasks, ranging from logical bugs to security vulnerabilities. While these generation failures are often treated as model-level limitations, empirical evidence increasingly traces their root causes to imperfections within the training corpora. Yet, the specific mechanisms linking training data quality issues to generated code quality issues remain largely unmapped. This paper presents a systematic literature review of 114 primary studies to investigate how training data quality issues propagate into code generation. We establish a unified taxonomy that categorizes generated code quality issues across nine dimensions and training data quality issues into code and non-code attributes. Based on this taxonomy, we formalize a causal framework detailing 18 typical propagation mapping mechanisms. Furthermore, we synthesize state-of-the-art detection and mitigation techniques across the data, model, and generation lifecycles. The reviewed literature reveals a clear methodological shift: quality assurance is transitioning from reactive, heuristic-based post-generation filtering toward proactive, data-centric governance and closed-loop repair. Finally, we identify open challenges and outline research directions for developing reliable LLMs for code through integrated data curation and continuous evaluation. Our repository is available at https://github.com/SYSUSELab/From-Data-to-Code.

cs.SE↗

arXiv2026

Large magnetoresistance and weak-antilocalization in the nodal-line semimetal VP2

After growing successfully high quality VP$_2$ single crystals, we studied systematically their longitudinal $\rho_{xx}(T)$ and Hall resistivity $\rho_{yx}(T)$ at various magnetic fields, combining the electronic band and Fermi surface (FS) calculations. Band calculations reveal that VP$_2$ is a type-II nodal-line semimetal, evidenced by the Hall resistivity measurements. It is found that the magnetoresistance (MR) at higher magnetic fields exhibits a linear behavior and does not show any sign of saturation, reaching 170\% at 40 K up to 9 T, which is determined by the intrinsic electronic structure and dominated by the Lorenz force, demonstrated by the resistivity anisotropy measurements and the numerical simulations. We also found that the existence of small amount magnetic impurities (V$^{4+}$, $S=1/2$, 2.24\%) results in Kondo effect emerging in $\rho_{xx}(T)$, the conductivity at lower magnetic fields exhibits a typical weak anti-localization (WAL) behavior. These results illustrate that VP$_2$ is a platform to study the electronic transport properties of a topological material containing magnetic impurities.

cond-mat.str-el↗