Searcharxiv⌕ Search

arXiv subjects

Weifeng Sun

Publications and source records attributed to Weifeng Sun.

At least 19 recordsLinked to original sources
arXiv2026

A Seifert Dichotomy for Z2-Harmonic One-Forms

We establish a dichotomy for Z/2-harmonic one-forms on Seifert fibered rational homology three-spheres. For an oriented base with at most three exceptional fibers, we prove nonexistence along any fixed connection-metric family when the fibers are sufficiently collapsed. With at least four exceptional fibers, existing results give existence for every Riemannian metric. We also complete the corresponding classification for nonorientable bases.

math.DG↗
arXiv2026

How Reasoning Shapes Social Bias in LLM-Generated Code?

Large language models (LLMs) are increasingly used for code generation, yet generated programs may exhibit social bias through unfair or differential treatment of sensitive demographic attributes. While prior work mainly studies direct code generation, bias in reasoning-based generation remains underexplored. We conduct the first systematic study of social bias in reasoning-based code generation, evaluating 9 standard LLMs and large reasoning models (LRMs) on realistic bias-sensitive tasks across three human-centered decision scenarios. We find that reasoning generally reduces bias, lowering the average bias rate from 0.64 to 0.40, but the effect varies substantially across models. Meanwhile, code quality is not consistently preserved, with the average quality dropping from 0.72 to 0.59. Biased reasoning strongly predicts biased code, and adjusting generation configurations alone is insufficient for robust mitigation. Based on these findings, we propose ProbeDebias, a reasoning-aware framework that detects and rewrites biased reasoning traces before code generation. ProbeDebias achieves 87.76% F1 for reasoning-bias detection and reduces code bias by 83.73% on average while largely preserving quality. Compared with SOTA baselines, it further reduces average bias by 52.70%-54.42% and improves quality by 9.79%-36.79%. These results highlight the value of reasoning-stage analysis for trustworthy code generation.

cs.SE↗
arXiv2026

Understanding and Improving Model Editing for Secure Code Generation

Large language models (LLMs) are widely used for code generation, yet they can reproduce vulnerable implementations learned from insecure training patterns. Prior work has mainly explored inference-time hardening, which reduces insecure generations without modifying the target model but relies on auxiliary components and adds runtime overhead. We conduct the first systematic study of model editing as a model-level hardening mechanism for secure code generation. We evaluate 3 state-of-the-art editing methods across diverse LLM families and compare them with CoSec, a representative inference-time approach, focusing on security, robustness, generalization, and functional correctness. Model editing yields larger security gains than CoSec on seen vulnerability types, improving security ratios by 15%-25% over vanilla models, with gains remaining stable under prompt perturbations. However, these improvements transfer unreliably to unseen vulnerabilities and can reduce functional correctness. To mitigate this trade-off, we propose SafeEdit, a post-edit refinement method combining functional tuning with edit-aware regularization. Across eight target LLMs, SafeEdit improves Pass@1 over UltraEdit by 11.73/13.70/15.50 percentage points at T=0.1/0.4/0.8 while largely preserving security. Compared with CoSec, it achieves relative security-ratio gains of 7.54%-12.04%. Additional evaluation on CodeGuard+ confirms improved joint secure-and-correct generation. SafeEdit and CoSec are also complementary, and their combination can further improve security while maintaining strong functional correctness. Overall, our results provide evidence-backed guidance for applying model editing to secure code generation.

cs.CR↗
arXiv2026

AgentChaos: Chaos Engineering for Agent Systems via Programmatic Fault Injection

Agent systems rely on LLM APIs for every response, but these APIs can return server errors, truncated responses, or corrupted content that propagates through downstream agents and causes task failure. Evaluating robustness under these faults is crucial for reliable deployment. Existing fault injection methods are offline, require source code modification, or cannot modify specific response fields. A comprehensive evaluation also requires a systematic fault taxonomy because different fault types affect downstream agents differently. We propose AgentChaos, a chaos engineering framework for controlled, runtime, non-intrusive LLM API fault injection. Since all agent systems access LLMs through the same HTTP interface, we inject faults at this shared layer without modifying source code. We define crash, omission, and value faults on content and tool call fields, intercept and modify LLM API responses at runtime, and verify whether each fault is triggered to filter untriggered tasks and avoid underestimating fault impact. Evaluations across agent systems, benchmarks, and backbone LLMs under 65 fault configurations show that all systems degrade under fault injection, with pass@1 dropping by up to 50 percentage points. The ranking is consistent across models, suggesting that robustness depends on system implementation rather than model capability. Existing fault diagnosis methods achieve below 53% accuracy on fault type and below 56% on fault step, leaving room for improvement. We further reveal practical findings for agent system developers.

cs.SE↗
arXiv2026

Lossless Tensor Compression as Program Synthesis

Model checkpoints are growing in both number and size, which makes archival, transfer, and deployment increasingly costly. General-purpose compressors can reduce storage requirements but ignore tensor structure, whereas existing tensor-specific compressors rely on fixed and format-specific pipelines. We present Brevis, which formulates lossless tensor compression as program synthesis. We design a typed domain-specific language (DSL) that captures recurring tensor structures, such as repeated regions and floating-point fields, through a set of reversible operators. Given a tensor, Brevis synthesizes a self-contained DSL program that reconstructs it bit-exactly. A checkpoint-specific production prior, learned from a small representative sample of tensors, guides a bounded A* search to synthesize compact programs, which can later be executed directly for bit-exact decompression. On 10 public checkpoints spanning language, audio, and image generation models, Brevis reduces 2.13 TB of checkpoint data to 1.41 TB, a 33.93% storage reduction. It produces archives up to 30.87% smaller than those of four general-purpose compressors, including zstd and gzip, and smaller archives than the tensor-specific compressors ZipNN and DFloat11. Under a practical concurrency configuration, Brevis achieves 3.60 GB/s compression and 6.61 GB/s decompression while preserving every source byte.

cs.SE↗
arXiv2026

MultiFixer: A Coordinator-Proposer Based Multi-Agent Framework For Fixing Multi-Hunk Bugs

Automated Program Repair (APR) has benefited greatly from Large Language Models (LLMs), but existing LLM-based APR methods still struggle with multi-hunk bugs that require coordinated changes across multiple locations. These bugs demand repository-level context understanding, repair-order scheduling, and effective hunk-level patch generation and selection. To address these challenges, we propose MultiFixer, a novel Coordinator-Proposer based multi-agent framework for multi-hunk repair. MultiFixer performs tool-augmented bug analysis, constructs fine-grained repair context, iteratively generates patches through a Coordinator-Proposer architecture, and applies two-stage patch refinement for syntactic and semantic correctness. We evaluate MultiFixer on 835 bugs from Defects4J and three vulnerability benchmarks. On Defects4J, MultiFixer fixes 326 bugs, including 62 multi-method and 27 multi-file bugs, and outperforms prior APR baselines in the reported comparisons with the same base model. Moreover, MultiFixer also fixes 46 multi-hunk bugs among 95 unique fixes. When combined with Claude-3.5-Sonnet, MultiFixer repairs 420 bugs, establishing a new state of the art on Defects4J. On VUL4J, MultiFixer repairs 24 real-world vulnerabilities, including 5 multi-hunk cases. On the multi-hunk subsets of SEC-bench and PatchEval, MultiFixer fixes 11 and 19 vulnerabilities, respectively, outperforming all compared baselines under GPT-3.5. These results demonstrate the effectiveness of MultiFixer for multi-hunk repair.

cs.SE↗
arXiv2026

SciCodePile: A 128GB Corpus and Executable Benchmark for Challenging Scientific Code Generation

Large language models (LLMs) excel at general-purpose code generation, yet how well they handle scientific code remains an open question. Existing datasets and benchmarks are limited in scale, domain coverage, or executable verification, leaving the true gap between current LLMs and reliable scientific code generators inadequately assessed. To address these limitations, we present SciCodePile, the largest scientific code corpus to date, constructed from 37,737 public repositories and collectively comprising 128GB of code that spans multiple computational science disciplines. From this corpus, we further curate an executable benchmark of 200 tasks, each equipped with a sandboxed execution environment and an automated test harness for functional verification. We evaluate 15 LLMs from both open-source and closed-source families on three tasks: prefix-to-suffix completion, fill-in-the-middle infilling, and executable code generation. Results show that scientific code generation remains highly challenging: The best CodeBLEU reaches only 38.13 and 38.37 on the two completion tasks, while the strongest model achieves just 12.30\% Pass@1 on the executable benchmark, underscoring how far current models remain from reliable scientific code generation. To demonstrate the training utility of SciCodePile, we further show that continued pretraining on our corpus improves CodeBLEU by $\times$2.84 on scientific code completion, and instruction tuning on our data improves Pass@1 by $\times$4.79 on the executable benchmark. All code and data are available at https://huggingface.co/SciCodePile.

cs.SE↗
arXiv2026

Insecure Coding Preferences in Long-Term Memory: Security Risks for LLM-based Code Generation

LLM-based systems increasingly incorporate long-term memory to improve cross-session continuity. However, once insecure coding preferences are stored, they may silently influence security-critical decisions in subsequent generations. In this study, we conduct the first systematic empirical study on the impact of insecure coding preferences stored in long-term memory on the security of LLM-based code generation. We evaluate four LLMs (ChatGPT, Gemini, Qwen, and Grok) across five programming languages (Python, C, C++, Go, and JavaScript). Our results show that insecure memories significantly increase the risk of generating vulnerable code by 2.7-50.3 percentage points (pp). Moreover, they create a 5.4-14.0 percentage-point risk-warning gap, where warning-rate increases lag behind vulnerability-rate increases. Further analysis reveals that insecure memories are difficult to overwrite through normal interactions and can broadly influence model outputs even when prompts are phrased differently. Finally, we evaluate three mitigation strategies: security-requirement appending and memory storage reduce vulnerability rates by 19.7-33.6 pp but may degrade functional correctness by up to 15.9 pp; memory-level safety filtering achieves a 100\% detection rate on our evaluated risky memory entries and restores generation behavior to the without-memory baseline. Based on these findings, we provide actionable suggestions to improve the security of long-term memory in LLM-based code generation.

cs.CR↗
arXiv2026

ReProAgent: Tool-Augmented Multi-Stage Agentic Generation of Bug Reproduction Tests from Issue Reports

Reproduction tests help developers confirm reported issues and provide executable feedback for issue resolution, yet issue reports in open-source projects rarely include such tests. Recent studies have explored generating issue reproduction tests from issue reports with large language models, but existing approaches largely rely on prompt-based pipelines that retrieve textual context and generate tests. This limits their ability to understand how reported issues behave in repository-scale codebases and to flexibly organize the construction of reproduction tests. In this paper, we propose ReProAgent, a multi-stage agent framework for reproduction test generation from issue reports. ReProAgent decomposes the task into four agent stages: bug localization, root cause analysis, test planning, and test generation. To support these stages, ReProAgent integrates task-specific tools for task decomposition and reflection, context retrieval from both textual sources and repository graphs, and runtime interaction with the execution environment. Experiments on SWT-bench-lite and SWT-bench-verified show that ReProAgent successfully reproduces 58.43% and 70.30% of issues, outperforming all baselines, with an average cost of $0.14 per instance. For example, when equipped with GPT-5-mini, ReProAgent exceeds OpenHands with the same backbone by 20.43 and 7.90 percentage points, respectively. ReProAgent also generalizes across multiple backbone LLMs and improves downstream issue resolution performance when integrated with existing repair approaches.

cs.SE↗
arXiv2025

DynaFix: Iterative Automated Program Repair Driven by Execution-Level Dynamic Information

Automated Program Repair (APR) aims to automatically generate correct patches for buggy programs. Recent approaches leveraging large language models (LLMs) have shown promise but face limitations. Most rely solely on static analysis, ignoring runtime behaviors. Some attempt to incorporate dynamic signals, but these are often restricted to training or fine-tuning, or injected only once into the repair prompt, without iterative use. This fails to fully capture program execution. Current iterative repair frameworks typically rely on coarse-grained feedback, such as pass/fail results or exception types, and do not leverage fine-grained execution-level information effectively. As a result, models struggle to simulate human stepwise debugging, limiting their effectiveness in multi-step reasoning and complex bug repair. To address these challenges, we propose DynaFix, an execution-level dynamic information-driven APR method that iteratively leverages runtime information to refine the repair process. In each repair round, DynaFix captures execution-level dynamic information such as variable states, control-flow paths, and call stacks, transforming them into structured prompts to guide LLMs in generating candidate patches. If a patch fails validation, DynaFix re-executes the modified program to collect new execution information for the next attempt. This iterative loop incrementally improves patches based on updated feedback, similar to the stepwise debugging practices of human developers. We evaluate DynaFix on the Defects4J v1.2 and v2.0 benchmarks. DynaFix repairs 186 single-function bugs, a 10% improvement over state-of-the-art baselines, including 38 bugs previously unrepaired. It achieves correct patches within at most 35 attempts, reducing the patch search space by 70% compared with existing methods, thereby demonstrating both effectiveness and efficiency in repairing complex bugs.

cs.SE↗
arXiv2025

Intention Chain-of-Thought Prompting with Dynamic Routing for Code Generation

Large language models (LLMs) exhibit strong generative capabilities and have shown great potential in code generation. Existing chain-of-thought (CoT) prompting methods enhance model reasoning by eliciting intermediate steps, but suffer from two major limitations: First, their uniform application tends to induce overthinking on simple tasks. Second, they lack intention abstraction in code generation, such as explicitly modeling core algorithmic design and efficiency, leading models to focus on surface-level structures while neglecting the global problem objective. Inspired by the cognitive economy principle of engaging structured reasoning only when necessary to conserve cognitive resources, we propose RoutingGen, a novel difficulty-aware routing framework that dynamically adapts prompting strategies for code generation. For simple tasks, it adopts few-shot prompting; for more complex ones, it invokes a structured reasoning strategy, termed Intention Chain-of-Thought (ICoT), which we introduce to guide the model in capturing task intention, such as the core algorithmic logic and its time complexity. Experiments across three models and six standard code generation benchmarks show that RoutingGen achieves state-of-the-art performance in most settings, while reducing total token usage by 46.37% on average across settings. Furthermore, ICoT outperforms six existing prompting baselines on challenging benchmarks.

cs.AI↗
arXiv2025

PSearch: Search-based Patch Generation in the Era of LLM-based Automated Program Repair

Large Language Models (LLMs) have substantially advanced Automated Program Repair (APR), yet most existing LLM-based APR methods still rely on trial-and-error to generate patches. Such a strategy explores candidate patches in a weakly structured manner, making it difficult to assess the future potential of search directions and allocate search budget effectively. To address this limitation, we propose Psearch, a search-based patch generation framework for LLM-based APR centered on iterative patch evaluation and refinement. Instead of treating patch generation as repeated independent sampling, Psearch maintains a structured search state over intermediate patches, continuously evaluates the promise of explored search paths, and prioritizes the most promising ones for further refinement. This design enables Psearch to abandon weak directions early and progressively approach correct fixes through long-horizon search. Importantly, Psearch can be integrated with different search algorithms, while our current implementation adopts Monte Carlo Tree Search as one effective instantiation. We evaluate Psearch on five widely used bug and vulnerability benchmarks. Experimental results show that Psearch correctly repairs 201 out of 835 bugs in Defects4J, outperforming all 12 state-of-the-art baselines. Psearch also fixes 27 of 79 vulnerabilities in VUL4J and resolves 164 of 300 issues in SWE-Bench-Lite. Moreover, with a patch size of 16, Psearch reduces monetary cost to roughly 50% of strong baselines while maintaining superior repair effectiveness. These results highlight the effectiveness of Psearch for improving LLM-based APR. The code and results can be found at https://github.com/Tomsawyerhu/Psearch

cs.SE↗
arXiv2024

DFEPT: Data Flow Embedding for Enhancing Pre-Trained Model Based Vulnerability Detection

Software vulnerabilities represent one of the most pressing threats to computing systems. Identifying vulnerabilities in source code is crucial for protecting user privacy and reducing economic losses. Traditional static analysis tools rely on experts with knowledge in security to manually build rules for operation, a process that requires substantial time and manpower costs and also faces challenges in adapting to new vulnerabilities. The emergence of pre-trained code language models has provided a new solution for automated vulnerability detection. However, code pre-training models are typically based on token-level large-scale pre-training, which hampers their ability to effectively capture the structural and dependency relationships among code segments. In the context of software vulnerabilities, certain types of vulnerabilities are related to the dependency relationships within the code. Consequently, identifying and analyzing these vulnerability samples presents a significant challenge for pre-trained models. In this paper, we propose a data flow embedding technique to enhance the performance of pre-trained models in vulnerability detection tasks, named DFEPT, which provides effective vulnerability data flow information to pre-trained models. Specifically, we parse data flow graphs from function-level source code, and use the data type of the variable as the node characteristics of the DFG. By applying graph learning techniques, we embed the data flow graph and incorporate relative positional information into the graph embedding using sine positional encoding to ensure the completeness of vulnerability data flow information. Our research shows that DFEPT can provide effective vulnerability semantic information to pre-trained models, achieving an accuracy of 64.97% on the Devign dataset and an F1-Score of 47.9% on the Reveal dataset.

cs.SE↗
arXiv2023

Revisiting and Improving Retrieval-Augmented Deep Assertion Generation

Unit testing validates the correctness of the unit under test and has become an essential activity in software development process. A unit test consists of a test prefix that drives the unit under test into a particular state, and a test oracle (e.g., assertion), which specifies the behavior in that state. To reduce manual efforts in conducting unit testing, Yu et al. proposed an integrated approach (integration for short), combining information retrieval (IR) with a deep learning-based approach, to generate assertions for a unit test. Despite promising, there is still a knowledge gap as to why or where integration works or does not work. In this paper, we describe an in-depth analysis of the effectiveness of integration. Our analysis shows that: 1) The overall performance of integration is mainly due to its success in retrieving assertions. 2) integration struggles to understand the semantic differences between the retrieved focal-test (focal-test includes a test prefix and a unit under test) and the input focal-test; 3) integration is limited to specific types of edit operations and cannot handle token addition or deletion. To improve the effectiveness of assertion generation, this paper proposes a novel retrieve-and-edit approach named EditAS. Specifically, EditAS first retrieves a similar focal-test from a pre-defined corpus and treats its assertion as a prototype. Then, EditAS reuses the information in the prototype and edits the prototype automatically. EditAS is more generalizable than integration. We conduct experiments on two large-scale datasets and experimental results demonstrate that EditAS outperforms the state-of-the-art approaches, with an average improvement of 10.00%-87.48% and 3.30%-42.65% in accuracy and BLEU score, respectively.

cs.SE↗
arXiv2023

The extended Bogomolny equations on $R^2 \times R^+$ with real symmetry breaking

In this paper, we construct solutions to the extended Bogomolny equations on $X = R^2 \times R^+$ with certain boundary conditions and asymptotic conditions. Let $y$ be the coordinate of $R^+$. Roughly, both the boundary condition and the asymptotic condition say that a solution approaches to a certain model solution when $y \rightarrow 0$ and $y \rightarrow \infty$ resepctively. The boundary condition ($y \rightarrow 0$) is called generalized Nahm pole boundary condition and the asymptotic condition ($y \rightarrow \infty$) is called real symmetry breaking condition. The solutions should be thought as an analog of the instanton solutions that Taubes and Dimakis have created (using different methods), while their solutions satisfy a different asymptotic condition.

math.DG↗
arXiv2023

Toward Cost-effective Adaptive Random Testing: An Approximate Nearest Neighbor Approach

Adaptive Random Testing (ART) enhances the testing effectiveness (including fault-detection capability) of Random Testing (RT) by increasing the diversity of the random test cases throughout the input domain. Many ART algorithms have been investigated such as Fixed-Size-Candidate-Set ART (FSCS) and Restricted Random Testing (RRT), and have been widely used in many practical applications. Despite its popularity, ART suffers from the problem of high computational costs during test-case generation, especially as the number of test cases increases. Although several strategies have been proposed to enhance the ART testing efficiency, such as the forgetting strategy and the k-dimensional tree strategy, these algorithms still face some challenges, including: (1) Although these algorithms can reduce the computation time, their execution costs are still very high, especially when the number of test cases is large; and (2) To achieve low computational costs, they may sacrifice some fault-detection capability. In this paper, we propose an approach based on Approximate Nearest Neighbors (ANNs), called Locality-Sensitive Hashing ART (LSH-ART). When calculating distances among different test inputs, LSH-ART identifies the approximate (not necessarily exact) nearest neighbors for candidates in an efficient way. LSH-ART attempts to balance ART testing effectiveness and efficiency.

cs.SE↗
arXiv2022

ISTA-Inspired Network for Image Super-Resolution

Deep learning for image super-resolution (SR) has been investigated by numerous researchers in recent years. Most of the works concentrate on effective block designs and improve the network representation but lack interpretation. There are also iterative optimization-inspired networks for image SR, which take the solution step as a whole without giving an explicit optimization step. This paper proposes an unfolding iterative shrinkage thresholding algorithm (ISTA) inspired network for interpretable image SR. Specifically, we analyze the problem of image SR and propose a solution based on the ISTA method. Inspired by the mathematical analysis, the ISTA block is developed to conduct the optimization in an end-to-end manner. To make the exploration more effective, a multi-scale exploitation block and multi-scale attention mechanism are devised to build the ISTA block. Experimental results show the proposed ISTA-inspired restoration network (ISTAR) achieves competitive or better performances than other optimization-inspired works with fewer parameters and lower computation complexity.

eess.IV↗