arXiv subjects

Elliott Wen

Publications and source records attributed to Elliott Wen.

13 recordsLinked to original sources

arXiv2026

Can LLMs Produce Better Object-Oriented Designs than Human-Involved Development?

Background: Large Language Models (LLMs) are increasingly used for code generation. However, their ability to generate multi-class projects that require object-oriented design (OOD) remains unclear, especially relative to projects developed with human involvement. Aims: The primary objective of this study is to compare OOD quality in projects from three authorship conditions: PreAI (human-involved projects produced before widespread LLM use), PostAI (human-involved projects produced after widespread LLM use), and PureAI (projects generated end-to-end by contemporary LLMs). Method: We conducted a comparative case study on a postgraduate Java assignment. Two offerings of the same assignment were selected as the PreAI and PostAI datasets. PureAI projects were generated using three contemporary LLMs. We analyzed OOD quality using project-level OOD metrics, code smell density, and domain modeling. Results: Relative to human-involved projects, PureAI projects show lower code smell density and generally appear simpler in terms of total size, complexity, and coupling. However, this is consistent with oversimplification, as it is associated with missing abstractions and weaker responsibility separation. PostAI is closer to PureAI than PreAI on many OOD measures and also shows tendencies toward oversimplification. Conclusions: Our findings indicate that appropriate human guidance on object-oriented decomposition and responsibility assignment remains important when LLMs are used for object-oriented design.

cs.SE↗

arXiv2026

The Influence of Code Smells in Efferent Neighbors on Class Stability

Understanding what drives code instability is essential for effective software maintenance, as unstable classes require larger or more frequent edits and increase the risk of unintended side effects. Although code smells are widely believed to harm maintainability, most prior stability studies examine only the smells within the class being modified. In practice, however, classes can change because their efferent neighbors (i.e., the classes they depend on) are modified due to ripple effects that propagate along static dependencies, even if the class itself is clean. Such ripple effects may be more severe when the efferent neighbor exhibits code smells. In addition, code smells rarely occur alone. They often appear together within a class or across classes connected by static dependencies, a phenomenon known as code smell interrelation. Such interrelation can lead to code smell interaction, where smells are directly connected through static dependencies and may further compound maintainability issues. However, the effect of code smell interrelation and interaction on code quality remains largely underexplored. Therefore, this study investigates whether the presence of code smells in a class's efferent neighbors affects its stability, considering the factor of code smell interrelation and interaction. To achieve this, we mine one year of commit history from 100 top-starred GitHub projects, detect code smells and static dependencies, determine code smell interrelation and interaction, and model these factors as predictors of class stability.

cs.SE↗

arXiv2026

Playsemble: Learning Low-Level Programming Through Interactive Games

Teaching assembly programming is a fundamental component of undergraduate computer science education, yet many students struggle with its abstract and low-level concepts. Existing learning tools, such as simulators and visualisers, support understanding by exposing machine states. However, they often limit students to passive observation and provide few opportunities for meaningful interaction. To address these limitations, we introduce Playsemble, a gamified learning system that transforms assembly instructions into interactive, game-like tasks in which students control Pac-Man to collect items, avoid ghosts, and reach targets. Playsemble integrates a code editor, a CPU emulator, and visual debugging tools within a browser-based environment, allowing students to work offline without installation or configuration. It also provides immediate formative feedback enhanced by large language models. We deployed Playsemble in an undergraduate computer architecture course with 107 students. The course featured a sequence of assignments of increasing complexity, covering core concepts such as register and memory manipulation, control structures including loops and conditionals, and arithmetic operations. Our findings suggest that Playsemble promotes active experimentation, sustained engagement, and deeper conceptual understanding through meaningful game-based learning experiences.

cs.CY↗

arXiv2026

Rethinking Perplexity: Revealing the Impact of Input Length on Perplexity Evaluation in LLMs

Perplexity is a widely adopted metric for assessing the predictive quality of large language models (LLMs) and often serves as a reference metric for downstream evaluations. However, recent evidence shows that perplexity can be unreliable, especially when irrelevant long inputs are used, raising concerns for both benchmarking and system deployment. While prior efforts have employed selective input filtering and curated datasets, the impact of input length on perplexity has not been systematically studied from a systems perspective and input length has rarely been treated as a first-class system variable affecting both fairness and efficiency. In this work, we close this gap by introducing LengthBenchmark, a system-conscious evaluation framework that explicitly integrates input length, evaluation protocol design, and system-level costs, evaluating representative LLMs under two scoring protocols (direct accumulation and fixed window sliding) across varying context lengths. Unlike prior work that focuses solely on accuracy-oriented metrics, LengthBenchmark additionally measures latency, memory footprint, and evaluation cost, thereby linking predictive metrics to deployment realities. We further incorporate quantized variants not as a main contribution, but as robustness checks, showing that length-induced biases persist across both full-precision and compressed models. This design disentangles the effects of evaluation logic, quantization, and input length, and demonstrates that length bias is a general phenomenon that undermines fair cross-model comparison. Our analysis yields two key observations: (i) sliding window evaluation consistently inflates performance on short inputs, and (ii) both full-precision and quantized models appear to realise gains as the evaluated segment length grows.

cs.LG↗

arXiv2025

Mind the Gap: Revealing Inconsistencies Across Heterogeneous AI Accelerators

While NVIDIA remains the dominant provider of AI accelerators within cloud data center, emerging vendors such as AMD, Intel, Mac, and Huawei offer cost-effective alternatives with claims of compatibility and performance. This paper presents the first empirical study investigating divergence in machine learning model across heterogeneous AI accelerators. Utilizing an automated pipeline, we synthesize over 100,000 variant models derived from 4,000 real-world models and execute them across five different enterprise-grade accelerators. Our findings suggest that newer AI platforms from Mac and Huawei support at least 17\% fewer operators than NVIDIA. These platforms also exhibit a higher rate of output discrepancies (exceeding 5\%), which stem from differences in operator implementations, handling of exceptional numerical values, and instruction scheduling. They are also more susceptible to failures during model compilation-based acceleration, and in some cases, the compiled models produce outputs that differ noticeably from those generated using the standard execution mode. In addition, we identify 7 implementation flaws in PyTorch and 40 platform-specific issues across vendors. These results underscore the challenges of achieving consistent machine learning behavior in an increasingly diverse hardware ecosystem.

cs.DC↗

arXiv2025

VRScout: Towards Real-Time, Autonomous Testing of Virtual Reality Games

Virtual Reality (VR) has rapidly become a mainstream platform for gaming and interactive experiences, yet ensuring the quality, safety, and appropriateness of VR content remains a pressing challenge. Traditional human-based quality assurance is labor-intensive and cannot scale with the industry's rapid growth. While automated testing has been applied to traditional 2D and 3D games, extending it to VR introduces unique difficulties due to high-dimensional sensory inputs and strict real-time performance requirements. We present VRScout, a deep learning-based agent capable of autonomously navigating VR environments and interacting with virtual objects in a human-like and real-time manner. VRScout learns from human demonstrations using an enhanced Action Chunking Transformer that predicts multi-step action sequences. This enables our agent to capture higher-level strategies and generalize across diverse environments. To balance responsiveness and precision, we introduce a dynamically adjustable sliding horizon that adapts the agent's temporal context at runtime. We evaluate VRScout on commercial VR titles and show that it achieves expert-level performance with only limited training data, while maintaining real-time inference at 60 FPS on consumer-grade hardware. These results position VRScout as a practical and scalable framework for automated VR game testing, with direct applications in both quality assurance and safety auditing.

cs.LG↗

arXiv2025

Analyzing Dependency Distribution Changes Arising from Code Smell Interactions

Dependencies between modules can trigger ripple effects when changes are made, making maintenance complex and costly, so minimizing these dependencies is crucial. Consequently, understanding what drives dependencies is important. One potential factor is code smells, which are symptoms in code that indicate design issues and reduce code quality. When multiple code smells interact through static dependencies, their combined impact on quality can be even more severe. While individual code smells have been widely studied, the influence of their interactions remains underexplored. In this study, we aim to investigate whether and how the distribution of static dependencies changes in the presence of code smell interactions. We conducted a dependency analysis on 116 open-source Java systems to quantify these interactions by comparing cases where code smell interactions exist and where they do not. Our results suggest that overall, code smell interactions are linked to a significant increase in total dependencies in 28 out of 36 cases, and that all code smells are associated with a consistent change direction (increase or decrease) in certain dependency types when interacting with other code smells. Consequently, this information can be used to support more accurate code smell detection and prioritization, as well as to develop more effective refactoring strategies.

cs.SE↗

arXiv2024

Keep Me Updated: An Empirical Study of Proprietary Vendor Blobs in Android Firmware

Despite extensive security research on various Android components, such as kernel or runtime, little attention has been paid to the proprietary vendor blobs within Android firmware. In this paper, we conduct a large-scale empirical study to understand the update patterns and assess the security implications of vendor blobs. We specifically focus on GPU blobs because they are loaded into every process for displaying graphics user interfaces and can affect the entire system's security. We examine over 13,000 Android firmware releases between January 2018 and April 2024. Our results reveal that device manufacturers often neglect vendor blob updates. About 82\% of firmware releases contain outdated GPU blobs (up to 1,281 days). A significant number of blobs also rely on obsolete LLVM core libraries released more than 15 years ago. To analyze their security implications, we develop a performant fuzzer that requires no physical access to mobile devices. We discover 289 security and behavioral bugs within the blobs. We also present a case study demonstrating how these vulnerabilities can be exploited via WebGL. This work underscores the critical security concerns associated with vulnerable vendor blobs and emphasizes the urgent need for timely updates from device manufacturers.

cs.SE↗

arXiv2023

VR.net: A Real-world Dataset for Virtual Reality Motion Sickness Research

Researchers have used machine learning approaches to identify motion sickness in VR experience. These approaches demand an accurately-labeled, real-world, and diverse dataset for high accuracy and generalizability. As a starting point to address this need, we introduce `VR.net', a dataset offering approximately 12-hour gameplay videos from ten real-world games in 10 diverse genres. For each video frame, a rich set of motion sickness-related labels, such as camera/object movement, depth field, and motion flow, are accurately assigned. Building such a dataset is challenging since manual labeling would require an infeasible amount of time. Instead, we utilize a tool to automatically and precisely extract ground truth data from 3D engines' rendering pipelines without accessing VR games' source code. We illustrate the utility of VR.net through several applications, such as risk factor detection and sickness level prediction. We continuously expand VR.net and envision its next version offering 10X more data than the current form. We believe that the scale, accuracy, and diversity of VR.net can offer unparalleled opportunities for VR motion sickness research and beyond.

cs.AI↗

arXiv2023

Striving for Authentic and Sustained Technology Use In the Classroom: Lessons Learned from a Longitudinal Evaluation of a Sensor-based Science Education Platform

Technology integration in educational settings has led to the development of novel sensor-based tools that enable students to measure and interact with their environment. Although reports from using such tools can be positive, evaluations are often conducted under controlled conditions and short timeframes. There is a need for longitudinal data collected in realistic classroom settings. However, sustained and authentic classroom use requires technology platforms to be seen by teachers as both easy to use and of value. We describe our development of a sensor-based platform to support science teaching that followed a 14-month user-centered design process. We share insights from this design and development approach, and report findings from a 6-month large-scale evaluation involving 35 schools and 1245 students. We share lessons learnt, including that technology integration is not an educational goal per se and that technology should be a transparent tool to enable students to achieve their learning goals.

cs.HC↗

arXiv2022

Improving the Domain Adaptation of Retrieval Augmented Generation (RAG) Models for Open Domain Question Answering

Retrieval Augment Generation (RAG) is a recent advancement in Open-Domain Question Answering (ODQA). RAG has only been trained and explored with a Wikipedia-based external knowledge base and is not optimized for use in other specialized domains such as healthcare and news. In this paper, we evaluate the impact of joint training of the retriever and generator components of RAG for the task of domain adaptation in ODQA. We propose \textit{RAG-end2end}, an extension to RAG, that can adapt to a domain-specific knowledge base by updating all components of the external knowledge base during training. In addition, we introduce an auxiliary training signal to inject more domain-specific knowledge. This auxiliary signal forces \textit{RAG-end2end} to reconstruct a given sentence by accessing the relevant information from the external knowledge base. Our novel contribution is unlike RAG, RAG-end2end does joint training of the retriever and generator for the end QA task and domain adaptation. We evaluate our approach with datasets from three domains: COVID-19, News, and Conversations, and achieve significant performance improvements compared to the original RAG model. Our work has been open-sourced through the Huggingface Transformers library, attesting to our work's credibility and technical consistency.

cs.CL↗

arXiv2021

Fine-tune the Entire RAG Architecture (including DPR retriever) for Question-Answering

In this paper, we illustrate how to fine-tune the entire Retrieval Augment Generation (RAG) architecture in an end-to-end manner. We highlighted the main engineering challenges that needed to be addressed to achieve this objective. We also compare how end-to-end RAG architecture outperforms the original RAG architecture for the task of question answering. We have open-sourced our implementation in the HuggingFace Transformers library.

cs.IR↗

arXiv2019

Optimizing Controller Placement for Software-Defined Networks

Controller placement problem (CPP) is a key issue for Software-Defined Networking (SDN) with distributed controller architectures. This problem aims to determine a suitable number of controllers deployed in important locations so as to optimize the overall network performance. In comparison to communication delay, existing literature on the CPP assumes that the influence of controller workload distribution on network performance is negligible. In this paper, we tackle the CPP that simultaneously considers the communication delay, the control plane utilization, and the controller workload distribution. Due to this reason, our CPP is intrinsically different from and clearly more difficult than any previously studied CPPs that are NP-hard. To tackle this challenging issue, we develop a new algorithm that seamlessly integrates the genetic algorithm (GA) and the gradient descent (GD) optimization method. Particularly, GA is used to search for suitable CPP solutions. The quality of each solution is further evaluated through GD. Simulation results on two representative network scenarios (small-scale and large-scale) show that our algorithm can effectively strike the trade-off between the control plane utilization and the network response time.

cs.NI↗