Software Engineering Commons^™

Sustainable Llm Inference For Edge Ai: Evaluating Quantized Llms For Energy Efficiency, Output Accuracy, And Inference Latency, Erik Johanne Husom, Arda Goknil, Merve Astekin, Lwin Khin Shar, Andre Kasen, Sagar Sen, Benedikt Andreas Mithassel, Ahmet Soylu

Research Collection School Of Computing and Information Systems

Deploying Large Language Models (LLMs) on edge devices presents significant challenges due to computational constraints, memory limitations, inference speed, and energy consumption. Model quantization has emerged as a key technique to enable efficient LLM inference by reducing model size and computational overhead. In this study, we conduct a comprehensive analysis of 28 quantized LLMs from the Ollama library, which applies by default Post-Training Quantization (PTQ) and weight-only quantization techniques, deployed on an edge device (Raspberry Pi 4 with 4GB RAM). We evaluate energy efficiency, inference performance, and output accuracy across multiple quantization levels and task types. Models are benchmarked on …

Enhancing Spatial Understanding In Mixed-Reality Presentations, Nam-Dang Vo, Van-Vinh Thai, Nam-Hoi Do, Viet-Tham Huynh, Anthony Tang, Khan-Duy Le

Research Collection School Of Computing and Information Systems

Mixed reality (MR) presentations often involve a presenter wearing a head-mounted display (HMD) and an audience watching via a large display, making it difficult for audiences to perceive spatial relationships between the presenter and virtual objects. We report two experiments testing three design variations: (1) scene camera placement (audience-aligned vs. opposite), (2) overlaying the presenter’s first-person view, and (3) highlighting objects in the presenter’s view. Results show that audience-aligned cameras and object highlighting improve spatial understanding, while combining third- and first-person views can further aid perception. We derive design guidelines for configuring MR presentations to better support audience comprehension.

Designing For Novice Debuggers: A Pilot Study On An Ai-Assisted Debugging Tool, Oka Kurniawan, Erick Chandra, Christopher M. Poskitt, Yannic Noller, Kenny T.W. Choo, Cyrille Jegourel

Research Collection School Of Computing and Information Systems

Debugging is a fundamental skill that novice programmers must develop. Numerous tools have been created to assist novice programmers in this process. Recently, large language models (LLMs) have been integrated with automated program repair techniques to generate fixes for students' buggy code. However, many of these tools foster an over-reliance on AI and do not actively engage students in the debugging process. In this work, we aim to design an intuitive debugging assistant, CodeHinter, that combines traditional debugging tools with LLM-based techniques to help novice debuggers fix semantic errors while promoting active engagement in the debugging process. We present findings …

Simulated Interactive Debugging, Yannic Noller, Erick Chandra, Srinidhi Chandrashekar, Kenny Choo, Cyrille Jegourel, Oka Kurniawan, Christopher M. Poskitt

Research Collection School Of Computing and Information Systems

Debugging software, i.e., the localization of faults and their repair, is a key activity in software engineering. Therefore, effective and efficient debugging is one of the core skills a software engineer must develop. However, the teaching of debugging techniques is usually very limited or only taught in indirect ways, e.g., during software projects. As a result, most Computer Science (CS) students learn debugging only in an ad-hoc and unstructured way. In this work, we present our approach called Simulated Interactive Debugging that interactively guides students along the debugging process. The guidance aims to empower the students to repair their solutions …

Trustworthy Federated Learning Framework For Secure, Efficient, And Quality-Aware Distributed Ai, Asadullah Tariq

Dissertations

Federated Learning (FL) emerged as a significant advancement in the field of Artificial Intelligence (AI), enabling collaborative model training across distributed devices while maintaining data privacy. As the importance of FL and its application in various areas increased, addressing trustworthiness issues in its various aspects became crucial. In the FL process, clients contribute updates computed on their local datasets, which the server aggregates to iteratively refine the global model. However, not all client data may be relevant to the learning objective, and incorporating updates from irrelevant data can harm the model's performance. The selection of training samples significantly impacts model …

Exploring Autonomous Agents: A Closer Look At Why They Fail When Completing Tasks, Ruofan Lu, Yichen Li, Yintong Huo

Research Collection School Of Computing and Information Systems

Autonomous agent systems powered by Large Language Models (LLMs) have demonstrated promising capabilities in automating complex tasks. However, current evaluations largely rely on success rates without systematically analyzing the interactions, communication mechanisms, and failure causes within these systems. To bridge this gap, we present a benchmark of 34 representative programmable tasks designed to rigorously assess autonomous agents. Using this benchmark, we evaluate three popular open-source agent frameworks combined with two LLM backbones, observing a task completion rate of approximately 50%. Through in-depth failure analysis, we develop a three-tier taxonomy of failure causes aligned with task phases, highlighting planning errors, task …

Envisioning Future Interactive Web Development: Editing Webpage With Natural Language, Truong Hai Dang, Jingyu Xiao, Yintong Huo

Research Collection School Of Computing and Information Systems

The evolution of web applications relies on iterative code modifications, a process that is traditionally manual and time-consuming. While Large Language Models (LLMs) can generate UI code, their ability to edit existing code from new design requirements (e.g., ”center the logo”) remains a challenge. This is largely due to the absence of large-scale, high-quality tuning data to align model performance with human expectations. In this paper, we introduce a novel, automated data generation pipeline that uses LLMs to synthesize a high-quality fine-tuning dataset for web editing, named Instruct4Edit. Our approach generates diverse instructions, applies the corresponding code modifications, and performs …

Interaction2code: Benchmarking Mllm-Based Interactive Webpage Code Generation From Interactive Prototyping, Jingyu Xiao, Yuxuan Wan, Yintong Huo, Zixin Wang, Xinyi Xu, Wenxuan Wang, Zhiyao Xu, Yuhang Wang, Michael R. Lyu

Research Collection School Of Computing and Information Systems

Multimodal Large Language Models (MLLMs) have demonstrated remarkable performance on the design-to-code task, i.e., generating UI code from UI mock-ups. However, existing benchmarks only contain static web pages for evaluation and ignore the dynamic interaction, limiting the practicality, usability and user engagement of the generated webpages. To bridge these gaps, we present the first systematic investigation of MLLMs in generating interactive webpages. Specifically, we formulate the Interaction-to-Code task and establish the Interaction2Code benchmark, encompassing 127 unique webpages and 374 distinct interactions across 15 webpage types and 31 interaction categories. Through comprehensive experiments utilizing state-of-theart (SOTA) MLLMs, evaluated via both automatic …

Generative Ai And Empirical Software Engineering: A Paradigm Shift, Christoph Treude, Margaret-Anne Storey

Research Collection School Of Computing and Information Systems

The widespread adoption of generative AI in software engineering marks a paradigm shift, offering new opportunities to design and utilize software engineering tools while influencing both developers and the artifacts they create. Traditional empirical methods in software engineering, including quantitative, qualitative, and mixed-method approaches, are well established. However, this paradigm shift introduces novel data types and redefines many concepts in the software engineering process. The roles of developers, users, agents, and researchers increasingly overlap, blurring the distinctions between these social and technical actors within the field. This paper examines how integrating AI into software engineering challenges traditional research paradigms. It …

Teaching Diffusion Models To Ground Alpha Matte, Tianyi Xiang, Weiying Zheng, Yutao Jiang, Tingrui Shen, Hewei Yu, Yangyang Xu, Shengfeng He

Research Collection School Of Computing and Information Systems

The power of visual language models is showcased in visual understanding tasks, where language-guided models achieve impressive flexibility and precision. In this paper, we ex tend this capability to the challenging domain of image matting by framing it as a soft grounding problem, enabling a single diffusion model to handle diverse objects, textures, and transparencies, all directed by descriptive text prompts. Our method teaches the diffusion model to ground alpha mattes by guiding it through a process of instance-level localization and transparency estimation. First, we introduce an intermediate objective that trains the model to accurately localize semantic components of the …

Reproducibility Debt In Scientific Software, Zara Hassan, Christoph Treude, Graham Williams, Michael Norrish, Alex Potanin

Research Collection School Of Computing and Information Systems

Reproducibility Debt (RpD) refers to accumulated technical and organisational issues in scientific software that hinder the ability to reproduce research results. While reproducibility is essential to scientific integrity, RpD remains poorly defined and under-addressed. This study introduces a formal definition of RpD and investigates its causes, effects, and mitigation strategies using a mixed-methods approach involving a systematic literature review (214 papers), interviews (23 practitioners), and a global survey (59 participants). We identify seven categories of contributing issues, 75 causes, 110 effects, and 61 mitigation strategies. Findings are synthesised into a cause-effect model and supported by taxonomies of team roles and …

Viewsrd: 3d Visual Grounding Via Structured Multi-View Decomposition, Ronggang Huang, Haoxin Yang, Yan Cai, Xuemiao Xu, Huaidong Zhang, Shengfeng He

Research Collection School Of Computing and Information Systems

3Dvisual grounding aims to identify and localize objects in a 3Dspacebasedontextualdescriptions. However, existing methods struggle with disentangling targets from anchors in complex multi-anchor queries and resolving inconsisten cies in spatial descriptions caused by perspective variations. To tackle these challenges, we propose ViewSRD, a frame work that formulates 3D visual grounding as a structured multi-view decomposition process. First, the Simple Rela tion Decoupling (SRD) module restructures complex multi anchor queries into a set of targeted single-anchor state ments, generating a structured set of perspective-aware de scriptions that clarify positional relationships. These de composed representations serve as the foundation for the Multi-view …

Emoshortcuts: Emotionally Expressive Body Augmentation For Social Mixed Reality Avatars, Hyuna Seo, Youngki Lee, Rajesh Krishna Balan, Thivya Kandappu

Research Collection School Of Computing and Information Systems

We present EmoShortcuts1, a novel social Mixed Reality (MR) framework that enhances emotional expression by dynamically augmenting avatar body gestures to reflect users’ emotional states. While social MR enables immersive remote interactions through avatars, conveying emotions remains challenging due to limitations in head-mounted display (HMD) tracking (e.g., missing lower-body movements, such as stomping or defensive postures), and users’ tendency to deprioritize nonverbal expressions during multitasking. EmoShortcuts addresses these challenges by introducing an augmentation framework that generates expressive body gestures even when users’ physical movements are restricted. We conducted a formative study with 12 participants to identify key challenges in emotional …

Morphology-Aware Hrv Estimation From Wrist Ppg In Sedentary Scenarios, Changshuo Hu, Hung Manh Pham, Dong Ma

Research Collection School Of Computing and Information Systems

Photoplethysmography (PPG) is widely used in wearable devices for non-invasive heart rate variability (HRV) monitoring. While most prior work focuses on mitigating motion artifacts, recent studies highlight that even subtle contact pressure variations can distort waveform morphology and lead to inaccurate HRV estimates. In this work, we propose a morphology-aware deep learning framework that conditions HRV estimation on beat-level waveform types. Our model jointly encodes the raw PPG waveform and a sequence of pressure-induced morphology labels using parallel encoders, integrates them via cross-attention, and predicts normal-to-normal (NN) intervals and beat count to support downstream HRV computation. Evaluated on the public …

Developing A Strong Cps Defender: An Evolutionary Approach, Qingyuan Hu, Christopher M. Poskitt, Jun Sun, Yuqi Chen

Research Collection School Of Computing and Information Systems

Cyber-physical systems (CPSs) are used extensively in critical infrastructure, underscoring the need for anomaly detection systems that are able to catch even the most motivated attackers. Traditional anomaly detection techniques typically do `one-off' training on datasets crafted by experts or generated by fuzzers, potentially limiting their ability to generalize to unseen and more subtle attack strategies. Stopping at this point misses a key opportunity: a defender can actively challenge the attacker to find more nuanced attacks, which in turn can lead to more effective detection capabilities. Building on this concept, we propose Evo-Defender, an evolutionary framework that iteratively strengthens CPS …

Enhancing Llm Code Generation: A Systematic Evaluation Of Multi-Agent Collaboration And Runtime Debugging For Improving Accuracy, Reliability, And Latency, Nazmus Ashrafi

Theses

The use of large language models (LLMs) for automated code generation has emerged as a significant focus within AI research. As these pretrained models continue to evolve, their ability to understand and generate complex code structures has opened up new possibilities for automating intricate programming tasks with greater accuracy. Although contemporary foundational models demonstrate promising results, researchers continue to explore optimal post-training strategies to enhance code quality. These include supervised fine-tuning, retrieval-augmented generation (RAG), debugging, and many others. In this thesis, I combine two such widely used post training approaches—namely (1) multi agent collaboration and (2) runtime execution of information-based …

Better Digital Contracts With Prosocial Friction-In-Design, Brett Frischmann, Moshe Y. Vardi

Faculty Publications

Contract law is supposed to enable people to reach genuine agreements and cooperate. If this ideal was ever a reality, the rise of mass market contracts and boil­erplate rendered it pure fiction. Modern consumer contracts are incomprehensible to most people. No one reads them anyway.

Digital contracting involves design features that amplify traditional boilerplate harms and create others. For example, digital contracting is too cheap; low marginal costs lead to overexpansion in scale and scope. To make matters worse, the loss of autonomy from repeat engagement with digital contracting systems is pernicious. People become increasingly predictable and programmable as digital …

Tactile Data Comics: Combining Step-By-Step Presentation Of Tactile Graphics With Verbal Narration For The Blind And Visually Impaired, Yang Jiao, Ruoting Sun, Rong Luo, Xiwen Yao, Xinran She, Kotaro Hara, Yuewen Zhang, Xinyi Fu

Research Collection School Of Computing and Information Systems

Tactile graphics on a refreshable display have proven effective in enabling visually impaired people to comprehend pictorial content. To further evaluate the effectiveness of refreshable tactile displays in blind education, we designed tactile data comics, a method that combines step-by-step presentation of tactile graphics with verbal narration. We conducted a user study with sixteen visually impaired students to compare tactile data comics against verbal-only and static tactile graphics. Our findings show that tactile data comics significantly improve participants’ comprehension and engagement during the learning experience. These empirical results suggest that the integration of refreshable tactile displays and tactile data comics …

Information-Bottleneck Driven Binary Neural Network For Change Detection, Kaijie Yin, Zhiyuan Zhang, Shu Kong, Tian Gao, Cheng-Zhong Xu, Hui Kong

Research Collection School Of Computing and Information Systems

In this paper, we propose Binarized Change Detection (BiCD), the first binary neural network (BNN) designed specifically for change detection. Conventional network binarization approaches, which directly quantize both weights and activations in change detection models, severely limit the network's ability to represent input data and distinguish between changed and unchanged regions. This results in significantly lower detection accuracy compared to real-valued networks. To overcome these challenges, BiCD enhances both the representational power and feature separability of BNNs, improving detection performance. Specifically, we introduce an auxiliary objective based on the Information Bottleneck (IB) principle, guiding the encoder to retain essential input …

From Release To Adoption: Challenges In Reusing Pre-Trained Ai Models For Downstream Developers, Peerachai Banyongrakkul, Mansooreh Zahedi, Patanamon Thongtanunam, Christoph Treude, Haoyu Gao

Research Collection School Of Computing and Information Systems

Pre-trained models (PTMs) have gained widespread popularity and achieved remarkable success across various fields, driven by their groundbreaking performance and easy accessibility through hosting providers. However, the challenges faced by downstream developers in reusing PTMs in software systems are less explored. To bridge this knowledge gap, we qualitatively created and analyzed a dataset of 840 PTM-related issue reports from 31 OSS GitHub projects. We systematically developed a comprehensive taxonomy of PTM-related challenges that developers face in downstream projects. Our study identifies seven key categories of challenges that downstream developers face in reusing PTMs, such as model usage, model performance, and …

Finding Safety Violations Of Ai-Enabled Control Systems Through The Lens Of Synthesized Proxy Programs, Jieke Shi, Zhou Yang, Junda He, Bowen Xu, Dongsun Kim, Donggyun Han, David Lo

Research Collection School Of Computing and Information Systems

Given the increasing adoption of modern AI-enabled control systems, ensuring their safety and reliability has become a critical task in software testing. One prevalent approach to testing control systems is falsification, which aims to find an input signal that causes the control system to violate a formal safety specification using optimization algorithms. However, applying falsification to AI-enabled control systems poses two significant challenges: (1) it requires the system to execute numerous candidate test inputs, which can be time-consuming, particularly for systems with AI models that have many parameters, and (2) multiple safety requirements are typically defined as a conjunctive specification, …

Boosting Symbolic Execution For Vulnerability Detection, Haoxin Tu

Dissertations and Theses Collection (Open Access)

Software systems written by humans tend to be unreliable and insecure, hence, bugs or vulnerabilities in them are inevitable. Symbolic execution has shown considerable potential in detecting diverse types of software bugs and also vulnerabilities that have severe security implications. However, existing symbolic execution engines still suffer from at least three fundamental limitations in memory modeling, path exploration, and structured input generation, which significantly impede existing engines from efficiently and effectively detecting software bugs and vulnerabilities.

The objective of this dissertation is to boost existing symbolic execution engines by designing a new memory model, two new path exploration strategies, and …

Apidocbooster: An Extract-Then-Abstract Framework Leveraging Large Language Models For Augmenting Api Documentation, Chengran Yang, Jiakun Liu, Bowen Xu, Christoph Treude, Yunbo Lyu, Junda He, Ming Li, David Lo

Research Collection School Of Computing and Information Systems

API documentation is often the most trusted resource for programming. Many approaches have been proposed to augment API documentation by summarizing complementary information from external resources such as Stack Overflow. Existing extractive-based summarization approaches excel in producing faithful summaries that accurately represent the source content without input length restrictions. Nevertheless, they suffer from inherent readability limitations. On the other hand, our empirical study on the abstractive-based summarization method, i.e., GPT-4, reveals that GPT-4 can generate coherent and concise summaries but presents limitations in terms of informativeness and faithfulness. We introduce APIDocBooster, an extract-then-abstract framework that seamlessly fuses the advantages of …

Static Analysis As A Feedback Loop: Enhancing Llm-Generated Code Beyond Correctness, Scott Blyth, Sherlock Licorish, Christoph Treude, Markus Wagner

Research Collection School Of Computing and Information Systems

Large language models (LLMs) have demonstrated impressive capabilities in code generation, achieving high scores on benchmarks such as HumanEval and MBPP. However, these benchmarks primarily assess functional correctness and neglect broader dimensions of code quality, including security, reliability, readability, and maintainability. In this work, we systematically evaluate the ability of LLMs to generate high-quality code across multiple dimensions using the PythonSecurityEval benchmark. We introduce an iterative static analysis-driven prompting algorithm that leverages Bandit and Pylint to identify and resolve code quality issues. Our experiments with GPT-4o show substantial improvements: security issues reduced from >40% to 13%, readability violations from >80% …

Educator Perceptions Of Devops Teaching Recommendations And Their Alignment With Common Challenges, Marcelo Romulo Fernandes, Pablo Paiva, Samuel Lucas De Moura Ferino, Roberta Coelho, Christoph Treude, Eduardo Aranha, Uirá Kulesza

Research Collection School Of Computing and Information Systems

DevOps education presents unique pedagogical challenges due to the diversity of tools, rapid technological change, and the multidisciplinary nature of the field. Although previous work has proposed recommendations to address these challenges, it is unclear how educators perceive these recommendations and whether they align with the challenges encountered in practice. In this paper, we present a quantitative and qualitative methods study involving 11 DevOps educators who interacted with Improve, a tool that presents a curated set of educational challenges and recommendations derived from previous literature. Educators indicated which recommendations they already use, which they intend to use, and which challenges …

Studying Satd In Drone Systems With Human-Ai Collaboration, Leevi Rantala, Lwin Khin Shar, Mäntylä Mika V., Wei Minn, Naing Tun Yan

Research Collection School Of Computing and Information Systems

Background: Self-Admitted Technical Debt (SATD) refers to sub-optimal solutions that developers acknowledge within the source code. SATD research originated on Java projects but is expanding to other domains. We focus on SATD in drones, which are used for various critical tasks.Aims: The primary objective is to investigate SATD in drone systems. The second aim is to explore the integration of AI and human collaboration for SATD labelling and classification.Method: We conducted a sample study of SATD comments in drone systems (14 open source, 4 SDKs) to analyse the quantity and types of SATD comments present. Our study incorporates collaboration between …

Stylegan-∞: Extending Stylegan To Arbitrary-Ratio Translation With Stylebook, Yihua Dai, Tianyi Xiang, Bailin Deng, Yong Du, Hongmin Cai, Jing Qin, Shengfeng He

Research Collection School Of Computing and Information Systems

Although pre-trained large-scale generative models StyleGAN series have proven to be effective in various editing and translation tasks, they are limited to pre-defined fixed aspect ratio. To overcome this limitation, we propose StyleGAN-∞, a model that enables pre-trained StyleGAN to perform arbitrary-ratio conditional synthesis. Our key insight is to distill the expressive StyleGAN features into a StyleBook, such that an arbitrary-ratio condition can be translated to other forms by properly assembling pre-defined StyleBook vectors. To learn and leverage the StyleBook, we employ a network with three distinct stages, each corresponding to StyleBook extraction, StyleBook correspondence learning, and arbitrary-ratio synthesis. Extensive …

Improving Co-Decoding Based Security Hardening Of Code Llms Leveraging Knowledge Distillation, Dong Li, Shanfu Shu, Meng Yan, Zhongxin Liu, Chao Liu, Xiaohong Zhang, David Lo

Research Collection School Of Computing and Information Systems

Large Language Models (LLMs) have been widely adopted by developers in software development. However, the massive pretraining code data is not rigorously filtered, allowing LLMs to learn unsafe coding patterns. Several prior studies have demonstrated that code LLMs tend to generate code with potential vulnerabilities. The widespread adoption of intelligent programming assistants poses a significant threat to the software development process. Existing approaches to mitigating this risk primarily involve constructing secure data that are free of vulnerabilities and then retraining or fine-tuning the models. However, such an effort is resource intensive and requires significant manual supervision. When the model parameters …

Exploring Parameter-Efficient Fine-Tuning Techniques For Code Generation With Large Language Models, Martin Weyssow, Xin Zhou, Kisub Kim, David Lo, Houari A. Sahraoui

Research Collection School Of Computing and Information Systems

Large language models (LLMs) demonstrate impressive capabilities to generate accurate code snippets given natural language intents in a zero-shot manner, i.e., without the need for specific fine-tuning. While prior studies have highlighted the advantages of fine-tuning LLMs, this process incurs high computational costs, making it impractical in resource-scarce environments, particularly for models with billions of parameters. To address these challenges, previous research explored in-context learning (ICL) and retrieval-augmented generation (RAG) as strategies to guide the LLM generative process with task-specific prompt examples. However, ICL and RAG introduce inconveniences, such as the need for designing contextually relevant prompts and the absence …

Ponzilens+: Visualizing Bytecode Actions For Smart Ponzi Scheme Identification, Xiaolin Wen, Tai D. Nguyen, Shaolun Ruan, Qiaomu Shen, Jun Sun, Feida Zhu, Yong Wang

Research Collection School Of Computing and Information Systems

With the prevalence of smart contracts, smart Ponzi schemes have become a common fraud on blockchain and have caused significant financial loss to cryptocurrency investors in the past few years. Despite the critical importance of detecting smart Ponzi schemes, a reliable and transparent identification approach adaptive to various smart Ponzi schemes is still missing. To fill the research gap, we first extract semantic-meaningful actions to represent the execution behaviors specified in smart contract bytecodes, which are derived from a literature review and in-depth interviews with domain experts. We then propose PonziLens+, a novel visual analytic approach that provides an intuitive …