Searcharxiv⌕ Search

arXiv subjects

Zhiqing Zhong

Publications and source records attributed to Zhiqing Zhong.

14 recordsLinked to original sources
arXiv2026

SpecGuardBench: Can Coding Agents Construct Reliable Validation Harnesses from Natural Language Specifications?

As coding agents become more capable, tests remain the primary evidence for judging whether their code is correct. Yet passing available tests does not establish that an implementation satisfies its natural language specification. Reliable validation requires checks that faithfully capture the specification and scenarios that thoroughly exercise the code, but current benchmarks provide limited insight into whether agents can construct such evidence themselves. We introduce SpecGuardBench, a benchmark of 291 tasks from 13 real-world repositories, covering 8 types of verification logic such as functional postconditions and interaction protocols. Given a natural language specification, repository context, and a reference implementation, an agent constructs a validation harness comprising executable checks and test scenarios. The checks should reject implementations that violate the specification while accepting those that satisfy it, and the scenarios should thoroughly exercise the code. Across four proprietary and open-weight coding agents, none constructs a harness scored as faithful on the benchmark's implementations for more than 63% of the tasks. The bottleneck lies less in exploring test scenarios than in constructing checks that faithfully reflect the specification. Specifically, the agents show a clear tendency to write checks that mirror the reference implementation rather than the specification, and therefore reject implementations that the specification permits.

cs.CL↗
arXiv2026

Don't Guess, Just Ask: Evaluating Interactive Requirement Clarification for Repository-Level Software Development

As LLMs become increasingly capable of implementing software, a critical bottleneck emerges: whether they truly understand what users want. Users often begin with vague intent, leaving important requirements to be clarified through interaction. In repository-level development, misunderstandings can propagate and cause costly rework. However, existing benchmarks typically evaluate implementation from complete specifications or underspecified requests, overlooking pre-coding requirement clarification. Prior clarification studies mainly focus on function-level tasks. It therefore remains unclear how well LLMs can transform vague intent into clear requirements through interaction and how this capability affects subsequent code generation. We introduce ReqClarifyBench, an interactive requirement clarification benchmark for repository-level software development. It contains 341 tasks spanning game, miniapp, and web development, each with a vague user intent, a human-validated requirement document, and code evaluation artifacts. We employ LLM-based user simulators with diverse personas derived from real-world conversations. We evaluate six LLMs in terms of clarification effectiveness and generated requirement document quality. Results show that stronger clarification effectiveness does not necessarily translate into higher document quality, and performance varies substantially across user personas. We also find that coding-agent harnesses do not further improve requirement document quality compared with their base LLMs. These findings offer insights for building LLM-based end-to-end software development systems.

cs.SE↗
arXiv2026

KV-RM: Regularizing KV-Cache Movement for Static-Graph LLM Serving

Static-graph LLM decoders provide predictable launches, fixed tensor shapes, and low submission overhead, but online decoding exposes highly irregular KV-cache behavior: request lengths differ, EOS events arrive asynchronously, and logical histories fragment over time. Dynamic runtimes recover flexibility through paged KV management and step-level scheduling, while static-graph executors often over-reserve memory and suffer burst-time latency outliers. This paper studies whether much of this variability can be absorbed below a fixed decode interface. We present KV-RM, a runtime design that regularizes KV-cache movement beneath a static-graph LLM decoder. KV-RM decouples logical KV histories from physical storage, tracks active KV state through a block pager, and materializes each decode step through a single committed descriptor. A merge-staged transport path coalesces non-contiguous KV mappings into a small number of large transfer groups before a fixed-shape attention kernel consumes them. Optional bounded far-history summaries can be enabled under the same interface, but the core design does not depend on them. On a 2-GPU NVIDIA A100 node, KV-RM improves mixed-length decoding throughput and tail latency relative to a static-graph baseline, reduces reserved KV memory across workload families, and removes severe burst-time latency spikes under production-trace replay. These results suggest that KV-cache movement, rather than kernel shape, can be an effective boundary for recovering runtime flexibility in static-graph LLM serving.

cs.AR↗
arXiv2026

An Executable Benchmarking Suite for Tool-Using Agents

Closed-loop tool-using agents are increasingly evaluated in executable web, code, and micro-task environments, but benchmark reports often conflate workloads, action-generating drivers, and the evidence admitted for systems-facing claims. We present an executable benchmarking suite that makes these objects explicit under a shared evidence-admission contract. The suite connects WebArena Verified, a SWE-Gym slice with SWE-bench-compatible verification, and MiniWoB++ through common workload adapters, task manifests, event schemas, replay/freeze policy, declared drivers, and reporting pipelines. In the canonical release, the gate separates paper-facing evidence from preflight, fixture, smoke, and diagnostic rows while preserving non-admitted artifacts for audit and onboarding. The admitted evidence records latency, invalid-action behavior, patch-generation cost, verifier metadata, replay bindings, and provenance under one auditable contract. The gate is decision-relevant rather than merely clerical: in a separate WebArena Verified controller study, clean-baseline and medium live-stressed evaluation select different fixed controller variants under the same workload and admission contract. The release is scoped as a benchmarking suite and admitted evidence, not a new agent policy, model leaderboard, backend comparison, or autonomous SWE-bench solver.

cs.SE↗
arXiv2026

SWE-ABS: Adversarial Benchmark Strengthening Exposes Inflated Success Rates on Test-based Benchmark

The SWE-Bench Verified leaderboard is approaching saturation, with the top system achieving 78.80%. However, we show that this performance is inflated. Our re-evaluation reveals that one in five "solved" patches from the top-30 agents are semantically incorrect, passing only because weak test suites fail to expose their errors. We present SWE-ABS, an adversarial framework that strengthens test suites through a two-stage pipeline: (1) coverage-driven augmentation using program slicing to target untested code regions, and (2) mutation-driven adversarial testing that synthesizes plausible but incorrect patches to expose semantic blind spots. On SWE-Bench Verified (500 instances), SWE-ABS strengthens 50.2% of instances, a 25.1x improvement over prior work, and rejects 19.71% of previously passing patches. As a result, the top agent's score decreases from 78.80% to 62.20%, leading to significant leaderboard reshuffling, with the previous top-ranked agent dropping to fifth place.

cs.SE↗
arXiv2025

DABench-LLM: Standardized and In-Depth Benchmarking of Post-Moore Dataflow AI Accelerators for LLMs

The exponential growth of large language models has outpaced the capabilities of traditional CPU and GPU architectures due to the slowdown of Moore's Law. Dataflow AI accelerators present a promising alternative; however, there remains a lack of in-depth performance analysis and standardized benchmarking methodologies for LLM training. We introduce DABench-LLM, the first benchmarking framework designed for evaluating LLM workloads on dataflow-based accelerators. By combining intra-chip performance profiling and inter-chip scalability analysis, DABench-LLM enables comprehensive evaluation across key metrics such as resource allocation, load balance, and resource efficiency. The framework helps researchers rapidly gain insights into underlying hardware and system behaviors, and provides guidance for performance optimizations. We validate DABench-LLM on three commodity dataflow accelerators, Cerebras WSE-2, SambaNova RDU, and Graphcore IPU. Our framework reveals performance bottlenecks and provides specific optimization strategies, demonstrating its generality and effectiveness across a diverse range of dataflow-based AI hardware platforms.

cs.AR↗
arXiv2025

BackportBench: A Multilingual Benchmark for Automated Backporting of Patches

Many modern software projects evolve rapidly to incorporate new features and security patches. It is important for users to update their dependencies to safer versions, but many still use older, vulnerable package versions because upgrading can be difficult and may break their existing codebase. Software developers can mitigate this problem by backporting security patches to older releases. However, manually backporting is time-consuming and error-prone. The effectiveness of existing automated backporting techniques on general software remains unclear since they typically target only code-hunk or function-level patch porting scenarios and are evaluated with imperfect metrics. To facilitate the development and evaluation of automated backporting techniques, we introduce BackportBench, the first comprehensive benchmark suite for patch backporting problem. BackportBench is a multilingual benchmark that contains 202 patch backporting problems from PyPI, Maven, and npm, each with executable Docker environments and relevant test cases. We evaluated existing patch porting methods and LLM-based techniques that have the potential to adapt to this task using BackportBench. The results show that the agentic method has outperformed traditional patch porting methods, especially on cases that require logical and structural changes. However, the performance varies across different programming languages. Based on the findings, we draw several implications for researchers and software practitioners in future work on automated backporting.

cs.SE↗
arXiv2025

An Efficient Gradient-Aware Error-Bounded Lossy Compressor for Federated Learning

Federated learning (FL) enables collaborative model training without exposing clients' private data, but its deployment is often constrained by the communication cost of transmitting gradients between clients and the central server, especially under system heterogeneity where low-bandwidth clients bottleneck overall performance. Lossy compression of gradient data can mitigate this overhead, and error-bounded lossy compression (EBLC) is particularly appealing for its fine-grained utility-compression tradeoff. However, existing EBLC methods (e.g., SZ), originally designed for smooth scientific data with strong spatial locality, rely on generic predictors such as Lorenzo and interpolation for entropy reduction to improve compression ratio. Gradient tensors, in contrast, exhibit low smoothness and weak spatial correlation, rendering these predictors ineffective and leading to poor compression ratios. To address this limitation, we propose an EBLC framework tailored for FL gradient data to achieve high compression ratios while preserving model accuracy. The core of it is an innovative prediction mechanism that exploits temporal correlations across FL training rounds and structural regularities within convolutional kernels to reduce residual entropy. The predictor is compatible with standard quantizers and entropy coders and comprises (1) a cross-round magnitude predictor based on a normalized exponential moving average, and (2) a sign predictor that leverages gradient oscillation and kernel-level sign consistency. Experiments show that this new EBLC yields up to 1.53x higher compression ratios than SZ3 with lower accuracy loss. Integrated into a real-world FL framework, APPFL, it reduces end-to-end communication time by 76.1%-96.2% under various constrained-bandwidth scenarios, demonstrating strong scalability for real-world FL deployments.

cs.LG↗
arXiv2025

SWE-Effi: Re-Evaluating Software AI Agent System Effectiveness Under Resource Constraints

The advancement of large language models (LLMs) and code agents has demonstrated significant potential to assist software engineering (SWE) tasks, such as autonomous issue resolution and feature addition. Existing AI for software engineering leaderboards (e.g., SWE-bench) focus solely on solution accuracy, ignoring the crucial factor of effectiveness in a resource-constrained world. This is a universal problem that also exists beyond software engineering tasks: any AI system should be more than correct - it must also be cost-effective. To address this gap, we introduce SWE-Effi, a set of new metrics to re-evaluate AI systems in terms of holistic effectiveness scores. We define effectiveness as the balance between the accuracy of outcome (e.g., issue resolve rate) and the resources consumed (e.g., token and time). In this paper, we specifically focus on the software engineering scenario by re-ranking popular AI systems for issue resolution on a subset of the SWE-bench benchmark using our new multi-dimensional metrics. We found that AI system's effectiveness depends not just on the scaffold itself, but on how well it integrates with the base model, which is key to achieving strong performance in a resource-efficient manner. We also identified systematic challenges such as the "token snowball" effect and, more significantly, a pattern of "expensive failures". In these cases, agents consume excessive resources while stuck on unsolvable tasks - an issue that not only limits practical deployment but also drives up the cost of failed rollouts during RL training. Lastly, we observed a clear trade-off between effectiveness under the token budget and effectiveness under the time budget, which plays a crucial role in managing project budgets and enabling scalable reinforcement learning, where fast responses are essential.

cs.SE↗
arXiv2025

Bi-Grid Reconstruction for Image Anomaly Detection

In image anomaly detection, significant advancements have been made using un- and self-supervised methods with datasets containing only normal samples. However, these approaches often struggle with fine-grained anomalies. This paper introduces \textbf{GRAD}: Bi-\textbf{G}rid \textbf{R}econstruction for Image \textbf{A}nomaly \textbf{D}etection, which employs two continuous grids to enhance anomaly detection from both normal and abnormal perspectives. In this work: 1) Grids as feature repositories that improve generalization and mitigate the Identical Shortcut (IS) issue; 2) An abnormal feature grid that refines normal feature boundaries, boosting detection of fine-grained defects; 3) The Feature Block Paste (FBP) module, which synthesizes various anomalies at the feature level for quick abnormal grid deployment. GRAD's robust representation capabilities also allow it to handle multiple classes with a single model. Evaluations on datasets like MVTecAD, VisA, and GoodsAD show significant performance improvements in fine-grained anomaly detection. GRAD excels in overall accuracy and in discerning subtle differences, demonstrating its superiority over existing methods.

cs.CV↗
arXiv2024

An Empirical Study on Package-Level Deprecation in Python Ecosystem

Open-source software (OSS) plays a crucial role in modern software development. Utilizing OSS code can greatly accelerate software development, reduce redundancy, and enhance reliability. Python, a widely adopted programming language, is renowned for its extensive and diverse third-party package ecosystem. However, a significant number of OSS packages within the Python ecosystem are in poor maintenance, leading to potential risks in functionality and security. Consequently, it is essential to establish a deprecation mechanism to assist package developers and users in managing packages effectively. To facilitate the establishment of the package-level deprecation mechanism, this paper presents a mixed-method empirical study, including data analysis and surveys. We investigate the current practices of announcing, receiving, and handling package-level deprecation in the Python ecosystem. We also assess the benefits of having deprecation announcements for inactively maintained packages. Furthermore, we investigate the challenges faced by package developers and users and their expectations for future deprecation practices. Our findings reveal that 75.4% of inactive package developers have no intention of releasing deprecation declarations for various reasons, while 89.5% of users express a desire to be notified about the deprecation, highlighting a gap between developers and users; in many cases, no alternative solutions are available when deprecation occurs, emphasizing the need to explore practical approaches that enable seamless package handover and require less maintenance effort. Our work aims to enhance the understanding of existing package-level deprecation patterns within the Python OSS realm and facilitate the development of deprecation practices for the Python community in the future.

cs.SE↗
arXiv2023

ROME: Testing Image Captioning Systems via Recursive Object Melting

Image captioning (IC) systems aim to generate a text description of the salient objects in an image. In recent years, IC systems have been increasingly integrated into our daily lives, such as assistance for visually-impaired people and description generation in Microsoft Powerpoint. However, even the cutting-edge IC systems (e.g., Microsoft Azure Cognitive Services) and algorithms (e.g., OFA) could produce erroneous captions, leading to incorrect captioning of important objects, misunderstanding, and threats to personal safety. The existing testing approaches either fail to handle the complex form of IC system output (i.e., sentences in natural language) or generate unnatural images as test cases. To address these problems, we introduce Recursive Object MElting (Rome), a novel metamorphic testing approach for validating IC systems. Different from existing approaches that generate test cases by inserting objects, which easily make the generated images unnatural, Rome melts (i.e., remove and inpaint) objects. Rome assumes that the object set in the caption of an image includes the object set in the caption of a generated image after object melting. Given an image, Rome can recursively remove its objects to generate different pairs of images. We use Rome to test one widely-adopted image captioning API and four state-of-the-art (SOTA) algorithms. The results show that the test cases generated by Rome look much more natural than the SOTA IC testing approach and they achieve comparable naturalness to the original images. Meanwhile, by generating test pairs using 226 seed images, Rome reports a total of 9,121 erroneous issues with high precision (86.47%-92.17%). In addition, we further utilize the test cases generated by Rome to retrain the Oscar, which improves its performance across multiple evaluation metrics.

cs.SE↗
arXiv2022

Automated Testing of Image Captioning Systems

Image captioning (IC) systems, which automatically generate a text description of the salient objects in an image (real or synthetic), have seen great progress over the past few years due to the development of deep neural networks. IC plays an indispensable role in human society, for example, labeling massive photos for scientific studies and assisting visually-impaired people in perceiving the world. However, even the top-notch IC systems, such as Microsoft Azure Cognitive Services and IBM Image Caption Generator, may return incorrect results, leading to the omission of important objects, deep misunderstanding, and threats to personal safety. To address this problem, we propose MetaIC, the \textit{first} metamorphic testing approach to validate IC systems. Our core idea is that the object names should exhibit directional changes after object insertion. Specifically, MetaIC (1) extracts objects from existing images to construct an object corpus; (2) inserts an object into an image via novel object resizing and location tuning algorithms; and (3) reports image pairs whose captions do not exhibit differences in an expected way. In our evaluation, we use MetaIC to test one widely-adopted image captioning API and five state-of-the-art (SOTA) image captioning models. Using 1,000 seeds, MetaIC successfully reports 16,825 erroneous issues with high precision (84.9\%-98.4\%). There are three kinds of errors: misclassification, omission, and incorrect quantity. We visualize the errors reported by MetaIC, which shows that flexible overlapping setting facilitates IC testing by increasing and diversifying the reported errors. In addition, MetaIC can be further generalized to detect label errors in the training dataset, which has successfully detected 151 incorrect labels in MS COCO Caption, a standard dataset in image captioning.

cs.SE↗