arXiv subjects

Tse-Hsun Chen

Publications and source records attributed to Tse-Hsun Chen.

At least 19 recordsLinked to original sources

arXiv2026

LLM4Log: A Systematic Review of Large Language Model-based Log Analysis

Software systems generate massive, evolving, semi-structured logs that are central to reliability engineering and AIOps, yet difficult to analyze at scale under drift and limited labels. Recent advances in pretrained Transformer models and instruction-tuned large language models (LLMs) have reshaped log analysis by enabling semantic generalization and cross-source evidence integration, but also introducing deployment risks such as context limits, latency and cost, privacy constraints, and hallucinations. This paper presents LLM4Log, a systematic review of LLM-based log analysis across the end-to-end pipeline, from upstream logging-statement generation and maintenance to log parsing/structuring and downstream tasks including anomaly detection, failure prediction, root cause analysis, and log summarization. Following a structured search and manual screening protocol, we completed literature collection in November 2025 and identified 145 unique papers across seven logging tasks. We organize the research area through a unified, task-driven taxonomy, summarize common design patterns (prompting/ICL, retrieval grounding, fine-tuning, tool/agent augmentation, and verification), and analyze evaluation practices, datasets, metrics, and reproducibility. Based on these cross-paper analyses, we summarize key lessons and open challenges for reliable real-world adoption. We emphasize robustness under drift and long-tail events, grounding and faithfulness for operator-facing outputs, and deployment-oriented designs with verifiable behavior.

cs.SE↗

arXiv2026

Reusing Past Repairs Through Hierarchical Trajectory Abstraction for Coding Agents

Although LLM-driven repair agents can tackle complex, repository-level issues, they treat every issue independently and discard the procedural knowledge accumulated from previous repairs. We introduce STAIR, a framework that converts historical repair trajectories into hierarchical, reusable plans that can be adapted to steer future repairs. Each past trajectory is transformed into a multi-level tree that ranges from fine-grained diagnostic actions to high-level repair strategies, encoding experience at several granularities. When a new issue arrives, STAIR selects relevant plan nodes from multiple abstraction levels, tailors them into executable, issue-specific plans, and supplies them to the agent through its prompt. On SWE-bench Verified, STAIR integrated with Lingxi reaches 81.2% Pass@1 using MiniMax M2.5 and 79.2% using GPT-5. The generated plans also generalize across agents: without any code change, they lift the Pass@1 of a structurally different agent, mini-SWE-agent v2, from 75.8% to 81.0%. Ablation experiments further show that mixing multiple abstraction levels surpasses any single level and that raw, unabstracted trajectories transfer substantially worse.

cs.SE↗

arXiv2026

Preventing Premature Commitment in Coding Agents with an Evidence-Conditioned Execution Layer

LLM-based coding agents often edit source code or submit patches before examining enough repository evidence to justify the change, a failure pattern we call premature commitment. We present ECLoop, an execution layer that interposes between the agent and the repository to enforce evidence-conditioned execution. For each task, ECLoop uses the issue description and repository structure to compile a set of conditions specifying what the agent should observe before each type of code modification or patch submission. During execution, ECLoop tracks which conditions the agent's runtime trajectory has satisfied and postpones any proposed action whose required conditions remain unmet. Evaluated on all 500 instances of SWE-bench Verified with two language models and two agent scaffolds, ECLoop raises Pass@1 by 4.8-11.8 percentage points without model retraining or scaffold changes. Ablation experiments show that each of ECLoop's three operations contributes distinct value and that structured evidence conditions outperform an equivalent natural-language summary. These gains come at no additional inference cost: by redirecting the agent before it pursues unsupported actions, ECLoop lowers average token consumption by up to 12.1%.

cs.SE↗

arXiv2026

IssueTrojanBench: Benchmarking AI Coding Agents Against Malicious Issue Requests

AI coding agents powered by LLMs are increasingly integrated into real-world software development, where they generate, edit, and execute code with autonomous access to local files and tools. Coding agents inherit security risks from both the LLM backbone, where adversarial prompts, poisoned training data, and backdoor triggers can cause models to emit insecure or attacker-chosen code, and their agentic architecture, where tool-using autonomy enables induced misuse of external APIs, data exfiltration, and persistent compromise of development environments. This paper presents a systematic evaluation of malicious issue requests against state-of-the-art coding agents (Cursor, Claude Code, and Codex Desktop), powered by two major model families (OpenAI GPT-5.3 Codex/GPT-5.4 and Anthropic Sonnet 4.6). Our novel benchmark IssueTrojanBench contains malicious issues that are constructed based on four novel attack categories (i.e., embedded as malicious instructions in issues), six delivery vectors (e.g., PDF, or issue comment), and further augmented by perturbations. Our results reveal critical vulnerabilities in the as-deployed modern coding agents, i.e., 66.5% of the malicious issues from IssueTrojanBench penetrate all the guardrails (agent- and LLM-level) of coding agents. Our further analysis shows that rejection is almost entirely from LLMs rather than the agent frameworks, with GPT models broadly vulnerable and Sonnet 4.6 exhibiting more selective, risk-aware blocking of high-impact actions. Our evaluation also highlights that the current agent-level defense strategy offers limited additional protection for coding agents. Our findings highlight the urgent need for stronger agent- and model-level safety mechanisms to protect AI coding agents.

cs.CR↗

arXiv2026

LogSemFuse: Semantic Evidence Fusion for Explainable Log Anomaly Detection

Log anomaly detection is critical for reliability monitoring and failure diagnosis in modern software systems. Existing model-based detectors provide useful anomaly signals, but they can still miss anomalous sessions and typically expose only scores or labels rather than the operational semantics behind a decision. This lack of semantic evidence limits their ability to explain why a session is anomalous, even when the final anomaly label is correct. The gap matters in practice because operators need to distinguish urgent failures from benign deviations and trace suspicious sessions back to concrete operational behavior. LLMs can recover richer log semantics, but using them as standalone detectors or repeatedly generating free-form explanations can be costly and difficult to reuse. We present LogSemFuse, an evidence-guided plug-in framework that enhances existing backbone detectors without replacing their original pipelines. LogSemFuse combines backbone predictions with reusable semantic evidence from local event patterns, LLM-based semantic reasoning, and cluster-derived executable rules to produce both anomaly decisions and evidence-based explanations. The resulting output reports the final label together with the semantic evidence that supports it, such as fired local patterns, triggered rules, and LLM rationale. We evaluate LogSemFuse on HDFS, BGL, and Liberty using DeepLog, LogAnomaly, LogBERT, and NeuralLog as backbones. LogSemFuse improves every non-perfect baseline, preserves the already perfect case, recovers 98.8% of backbone false negatives, and produces explanations preferred over direct LLM explanations in a human study. These gains require only modest and stable inference-time overhead, showing that semantic augmentation can improve detection effectiveness and interpretability without imposing large runtime costs.

cs.SE↗

arXiv2026

Agentic Software Engineering: Foundational Pillars and a Research Roadmap

Agentic Software Engineering (SE 3.0) represents a new era where intelligent agents are tasked not with simple code generation, but with achieving complex, goal-oriented SE objectives. To harness these new capabilities while ensuring trustworthiness, we must recognize a fundamental duality within the SE field in the Agentic SE era, comprising two symbiotic modalities: SE for Humans and SE for Agents. This duality demands a radical reimagining of the foundational pillars of SE (actors, processes, tools, and artifacts) which manifest differently across each modality. We propose two purpose-built workbenches to support this vision. The Agent Command Environment (ACE) serves as a command center where humans orchestrate and mentor agent teams, handling outputs such as Merge-Readiness Packs (MRPs) and Consultation Request Packs (CRPs). The Agent Execution Environment (AEE) is a digital workspace where agents perform tasks while invoking human expertise when facing ambiguity or complex trade-offs. This bi-directional partnership, which supports agent-initiated human callbacks and handovers, gives rise to new, structured engineering activities (i.e., processes) that redefine human-AI collaboration, elevating the practice from agentic coding to true agentic software engineering. This paper presents the Structured Agentic Software Engineering (SASE) vision, outlining several of the foundational pillars for the future of SE. The paper culminates in a research roadmap that identifies a few key challenges and opportunities while briefly discussing the resulting impact of this future on SE education. Our goal is not to offer a definitive solution, but to provide a conceptual scaffold with structured vocabulary to catalyze a community-wide dialogue, pushing the SE community to think beyond its classic, human-centric tenets toward a disciplined, scalable, and trustworthy agentic future.

cs.SE↗

arXiv2026

Beyond the GUI Paradigm: Do Mobile Agents Need the Phone Screen?

Recent advances in mobile agents are dominated by the GUI paradigm, in which agents perceive UI information and emit screen interactions. However, mobile platforms also expose a command-line interface (CLI) that provides direct access to device services and data. We argue CLI deserves first-class consideration alongside GUI. We evaluate three coding agents (Claude Code, Terminus-2, mini-swe-agent) across four model APIs on AndroidWorld and MobileWorld without any mobile-specific post-training, comparing against three reproducible GUI baselines (GUI-Owl-1.5-32B, MAI-UI, Qwen3-VL-32B). Claude Code (Opus 4.7) reaches 71.8\% and 51.9\%, outperforming every reproducible GUI baseline (69.3/68.1/57.8\% on AndroidWorld; 43.2/26.3/13.3\% on MobileWorld), while every other CLI configuration remains competitive. To establish the paradigm's ceiling, we provide oracle CLI solutions that reach 88.8\% on AndroidWorld (103/116 tasks CLI-solvable) and 86.3\% on MobileWorld (101/117 tasks CLI-solvable), indicating substantial room for future improvement. To cover everyday user intents beyond the GUI scope, we introduce the \textbf{CLI-Advantage Task Suite}, comprising 45 templates across five categories: bulk operations, multi-condition filtering, aggregation, cross-app workflows, and hidden device state. Every CLI agent outperforms every GUI baseline in all five categories, with substantially fewer steps per task (10.7 vs.\ 18.6). To support future research on mobile CLI agents, we will open-source agent implementations, oracle solutions, the CLI-Advantage suite, and evaluation infrastructure.

cs.SE↗

arXiv2026

FALAT: Tracing Failures in LLM Agent Trajectories via Dependency-Guided Search

LLM-based agents increasingly solve complex tasks through long trajectories involving reasoning steps, tool calls, and inter-agent communication. However, when these agents fail, it is often unclear which agent caused the failure and which step introduced the decisive error. This attribution problem is challenging because mistakes can propagate across the trajectory: later actions may appear incorrect, but only because they depend on an earlier corrupted state. Therefore, failure attribution cannot be treated as independent step-level classification. We propose FALAT, a diagnostic framework for failure attribution in LLM agent trajectories. FALAT frames attribution as a dependency-guided search problem. It first constructs an expectation of how the task should be solved and uses this expectation to identify suspicious regions in the trajectory. It then traces dependencies among decisions, tool outputs, and agent messages to distinguish error-introducing steps from steps that merely inherit or propagate prior mistakes. Finally, FALAT evaluates whether correcting a candidate step would be sufficient to recover the expected outcome, allowing it to identify both the responsible agent and the decisive failure step. We evaluate FALAT on the Who&When benchmark, which includes both algorithm-generated and hand-crafted multi-agent failure trajectories. The results show that FALAT consistently improves responsible-agent and decisive-step attribution. Its best configurations achieve 46.0% step-level accuracy on algorithm-generated trajectories and 29.1% on the more challenging hand-crafted trajectories, outperforming specialized attribution baselines and direct prompting with standalone LLMs. These findings suggest that dependency-aware reasoning is essential for reliable failure diagnosis in LLM agent systems.

cs.AI↗

arXiv2026

Static Program Slicing Using Language Models With Dataflow-Aware Pretraining and Constrained Decoding

Static program slicing is a fundamental software engineering technique for isolating code relevant to specific variables. While recent learning-based approaches using language models (LMs) show promise in automating slice prediction, they suffer from inaccurate dependency modeling and unconstrained generation, where LMs fail to capture precise data flow relations and produce slices containing hallucinated tokens and statements. To address these challenges, we propose Sliceformer, a novel approach that reformulates static program slicing as a sequence-to-sequence task using small language models such as CodeT5+. Sliceformer introduces two key innovations that directly target the identified limitations. First, to improve dependency modeling, we design dataflow-aware pretraining objectives that leverage data flow graphs (DFG) to teach models data dependencies through dataflow-preserving statement permutation and dataflow-aware span corruption. Second, to eliminate hallucination, we develop a constrained decoding mechanism that enforces both lexical and syntactic constraints. We evaluate Sliceformer on Java and Python program slicing benchmarks, demonstrating consistent improvements over state-of-the-art baselines with up to 22% gain in ExactMatch.

cs.SE↗

arXiv2026

When LLMs Lag Behind: Knowledge Conflicts from Evolving APIs in Code Generation

The rapid evolution of software libraries creates a significant challenge for Large Language Models (LLMs), whose static parametric knowledge often becomes stale post-training. While retrieval-augmented generation (RAG) is commonly used to provide up-to-date API specifications, "context-memory conflict" arises when external instructions contradict a model's internal parametric knowledge. This paper presents a systematic empirical study of LLM code generation under API evolution (e.g., API deprecation, API modification, and API addition), by constructing a benchmark of 270 real-world updates from eight Python libraries. We evaluate four LLM families of 11 models. Our results show that without comprehensive documentation, LLMs struggle to prioritize external context, averaging only 42.55% of generated code examples are executable in the target environment. While structured documentation and larger model scales improve LLMs' ability to update adoption, they do not fully resolve executability issues with a low 66.36% executable rate. In addition, reasoning-based strategies (e.g., Self-Reflection) significantly boost LLMs' performance with 11% improvement on executable rate. Our findings highlight the persistence of outdated patterns from LLMs, even when API update specifications are provided, and emphasize the need for evolution-aware benchmarks and techniques.

cs.SE↗

arXiv2026

CODEPROMPTZIP: Code-specific Prompt Compression for Retrieval-Augmented Generation in Coding Tasks with LMs

Retrieval-Augmented Generation (RAG) enhances coding tasks by incorporating retrieved code examples into prompts. However, lengthy prompts, often exceeding tens of thousands of tokens, introduce challenges related to limited context windows of language models (LMs) and high computational costs. Existing prompt compression techniques focus on natural language, lacking tailored solutions for code. To address the gap, we propose CodePromptZip, a framework that compresses code examples before integrating into RAG workflows. Our framework employs a type-aware, priority-driven strategy to construct training samples for training code compression model. By using program analysis, we identify token types (e.g., Identifier) and perform ablation analysis to rank their removal priorities based on their impact on task performance. We then train a small LM as the compressor on these samples, enabling flexible compression conditioned on specified ratios while minimizing performance degradation. Specially, the compressor is augmented with a copy mechanism, allowing tokens to be directly copied from the original code snippets. Evaluation results show that CodePromptZip surpasses SOTA entropy-based and distillation-based baselines, improving by 23.4%, 28.7%, and 8.7% over the best baseline for Assertion Generation, Bugs2Fix, and Code Suggestion, respectively.

cs.SE↗

arXiv2026

A Survey of Code Review Benchmarks and Evaluation Practices in Pre-LLM and LLM Era

Code review is a critical practice in modern software engineering, helping developers detect defects early, improve code quality, and facilitate knowledge sharing. With the rapid advancement of large language models (LLMs), a growing body of work has explored automated support for code review. However, progress in this area is hindered by the lack of a systematic understanding of existing benchmarks and evaluation practices. Current code review datasets are scattered, vary widely in design, and provide limited insight into what review capabilities are actually being assessed. In this paper, we present a comprehensive survey of code review benchmarks spanning both the Pre-LLM and LLM eras (2015--2025). We analyze 99 research papers (58 Pre-LLM era and 41 LLM era) and extract key metadata, including datasets, evaluation metrics, data sources, and target tasks. Based on this analysis, we propose a multi-level taxonomy that organizes code review research into five domains and 18 fine-grained tasks. Our study reveals a clear shift toward end-to-end generative peer review, increasing multilingual coverage, and a decline in standalone change understanding tasks. We further identify limitations of current benchmarks and outline future directions, including broader task coverage, dynamic runtime evaluation, and taxonomy-guided fine-grained assessment. This survey provides a structured foundation for developing more realistic and comprehensive benchmarks for LLM-based code review.

cs.SE↗

arXiv2025

Evaluating Software Process Models for Multi-Agent Class-Level Code Generation

Modern software systems require code that is not only functional but also maintainable and well-structured. Although Large Language Models (LLMs) are increasingly used to automate software development, most studies focus on isolated, single-agent function-level generation. This work examines how process structure and role specialization shape multi-agent LLM workflows for class-level code generation. We simulate a Waterfall-style development cycle covering Requirement, Design, Implementation, and Testing using three LLMs (GPT-4o-mini, DeepSeek-Chat, and Claude-3.5-Haiku) on 100 Python tasks from the ClassEval benchmark. Our findings show that multi-agent workflows reorganize, rather than consistently enhance, model performance. Waterfall-style collaboration produces cleaner and more maintainable code but often reduces functional correctness (-37.8\% for GPT-4o-mini and -39.8\% for DeepSeek-Chat), with Claude-3.5-Haiku as a notable exception (+9.5\%). Importantly, process constraints shift failure characteristics: structural issues such as missing code decrease, while semantic and validation errors become more frequent. Among all stages, Testing exerts the strongest influence by improving verification coverage but also introducing new reasoning failures, whereas Requirement and Design have comparatively modest effects. Overall, this study provides empirical evidence that software process structure fundamentally alters how LLMs reason, collaborate, and fail, revealing inherent trade-offs between rigid workflow discipline and flexible problem-solving in multi-agent code generation.

cs.SE↗

arXiv2025

Order Matters! An Empirical Study on Large Language Models' Input Order Bias in Software Fault Localization

Large Language Models (LLMs) show great promise in software engineering tasks like Fault Localization (FL) and Automatic Program Repair (APR). This study investigates the impact of input order and context size on LLM performance in FL, a crucial step for many downstream software engineering tasks. We test different orders for methods using Kendall Tau distances, including "perfect" (where ground truths come first) and "worst" (where ground truths come last), using two benchmarks that consist of both Java and Python projects. Our results indicate a significant bias in order; Top-1 FL accuracy in Java projects drops from 57% to 20%, while in Python projects, it decreases from 38% to approximately 3% when we reverse the code order. Breaking down inputs into smaller contexts helps reduce this bias, narrowing the performance gap in FL from 22% to 6% and then to just 1% on both benchmarks. We then investigated whether the bias in order was caused by data leakage by renaming the method names with more meaningful alternatives. Our findings indicated that the trend remained consistent, suggesting that the bias was not due to data leakage. We also look at ordering methods based on traditional FL techniques and metrics. Ordering using DepGraph's ranking achieves 48% Top-1 accuracy, which is better than more straightforward ordering approaches like CallGraphDFS. These findings underscore the importance of how we structure inputs, manage contexts, and choose ordering methods to improve LLM performance in FL and other software engineering tasks.

cs.SE↗

arXiv2025

SLICET5: Static Program Slicing using Language Models with Copy Mechanism and Constrained Decoding

Static program slicing is a fundamental technique in software engineering. Traditional static slicing tools rely on parsing complete source code, which limits their applicability to real-world scenarios where code snippets are incomplete or unparsable. While recent research developed learning-based approaches to predict slices, they face critical challenges: (1) Inaccurate dependency identification, where models fail to precisely capture data and control dependencies between code elements; and (2) Unconstrained generation, where models produce slices with extraneous or hallucinated tokens not present in the input, violating the structural integrity of slices. To address these challenges, we propose \ourtool, a novel slicing framework that reformulates static program slicing as a sequence-to-sequence task using lightweight language models (e.g., CodeT5+). Our approach incorporates two key innovations. First, we introduce a copy mechanism that enables the model to more accurately capture inter-element dependencies and directly copy relevant tokens from the input, improving both dependency reasoning and generation constraint. Second, we design a constrained decoding process with (a) lexical constraint, restricting outputs to input tokens only, and (b) syntactic constraint, leveraging Tree Similarity of Edit Distance (TSED) monotonicity to detect structurally invalid outputs and discard them. We evaluate \ourtool on CodeNet and LeetCode datasets and show it consistently outperforms state-of-the-art baselines, improving ExactMatch scores by up to 27\%. Furthermore, \ourtool demonstrates strong performance on incomplete code, highlighting its robustness and practical utility in real-world development environments.

cs.SE↗

arXiv2025

A Multi-Agent Approach to Fault Localization via Graph-Based Retrieval and Reflexion

Identifying and resolving software faults remains a challenging and resource-intensive process. Traditional fault localization techniques, such as Spectrum-Based Fault Localization (SBFL), leverage statistical analysis of test coverage but often suffer from limited accuracy. While learning-based approaches improve fault localization, they demand extensive training datasets and high computational resources. Recent advances in Large Language Models (LLMs) offer new opportunities by enhancing code understanding and reasoning. However, existing LLM-based fault localization techniques face significant challenges, including token limitations, performance degradation with long inputs, and scalability issues in complex software systems. To overcome these obstacles, we propose LLM4FL, a multi-agent fault localization framework that utilizes three specialized LLM agents. First, the Context Extraction Agent applies an order-sensitive segmentation strategy to partition large coverage data within the LLM's token limit, analyze failure context, and prioritize failure-related methods. The Debugger Agent then processes the extracted data, which employs graph-based retrieval-augmented code navigation to reason about failure causes and rank suspicious methods. Finally, the Reviewer Agent re-evaluates the identified faulty methods using verbal reinforcement learning, engaging in self-criticism and iterative refinement. Evaluated on the Defects4J (V2.0.0) benchmark, which includes 675 faults from 14 Java projects, LLM4FL achieves an 18.55\% improvement in Top-1 accuracy over AutoFL and 4.82\% over SoapFL. It outperforms supervised techniques such as DeepFL and Grace, all without requiring task-specific training. Furthermore, its coverage segmentation and prompt chaining strategies enhance performance, increasing Top-1 accuracy by up to 22\%.

cs.SE↗

arXiv2025

Evaluating the Effectiveness and Efficiency of Demonstration Retrievers in RAG for Coding Tasks

Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by integrating external knowledge bases, achieving state-of-the-art results in various coding tasks. The core of RAG is retrieving demonstration examples, which is essential to balance effectiveness (generation quality) and efficiency (retrieval time) for optimal performance. However, the high-dimensional nature of code representations and large knowledge bases often create efficiency bottlenecks, which are overlooked in previous research. This paper systematically evaluates the efficiency-effectiveness trade-off of retrievers across three coding tasks: Program Synthesis, Commit Message Generation, and Assertion Generation. We examined six retrievers: two sparse (BM25 and BM25L) and four dense retrievers, including one exhaustive dense retriever (SBERT's Semantic Search) and three approximate dense retrievers (ANNOY, LSH, and HNSW). Our findings show that while BM25 excels in effectiveness, it suffers in efficiency as the knowledge base grows beyond 1000 entries. In large-scale retrieval, efficiency differences become more pronounced, with approximate dense retrievers offering the greatest gains. For instance, in Commit Generation task, HNSW achieves a 44x speed up, while only with a 1.74% drop in RougeL compared with BM25. Our results also show that increasing the number of demonstrations in the prompt doesn't always improve the effectiveness and can increase latency and lead to incorrect outputs. Our findings provide valuable insights for practitioners aiming to build efficient and effective RAG systems for coding tasks.

cs.SE↗

arXiv2025

GUIWatcher: Automatically Detecting GUI Lags by Analyzing Mobile Application Screencasts

The Graphical User Interface (GUI) plays a central role in mobile applications, directly affecting usability and user satisfaction. Poor GUI performance, such as lag or unresponsiveness, can lead to negative user experience and decreased mobile application (app) ratings. In this paper, we present GUIWatcher, a framework designed to detect GUI lags by analyzing screencasts recorded during mobile app testing. GUIWatcher uses computer vision techniques to identify three types of lag-inducing frames (i.e., janky frames, long loading frames, and frozen frames) and prioritizes the most severe ones that significantly impact user experience. Our approach was evaluated using real-world mobile application tests, achieving high accuracy in detecting GUI lags in screencasts, with an average precision of 0.91 and recall of 0.96. The comprehensive bug reports generated from the lags detected by GUIWatcher help developers focus on the more critical issues and debug them efficiently. Additionally, GUIWatcher has been deployed in a real-world production environment, continuously monitoring app performance and successfully identifying critical GUI performance issues. By offering a practical solution for identifying and addressing GUI lags, GUIWatcher contributes to enhancing user satisfaction and the overall quality of mobile apps.

cs.SE↗