Search

The adoption of Large Language Models (LLMs) is reshaping software development as developers integrate these LLMs into their applications. In such applications, prompts serve as the primary means of interacting with LLMs. Despite the widespread use of LLM-integrated applications, there is limited understanding of how developers manage and evolve prompts. This study presents the first empirical analysis of prompt evolution in LLM-integrated software development. We analyzed 1,262 prompt changes across 243 GitHub repositories to investigate the patterns and frequencies of prompt changes, their relationship with code changes, documentation practices, and their impact on system behavior. Our findings show that developers primarily evolve prompts through additions and modifications, with most changes occurring during feature development. We identified key challenges in prompt engineering: only 21.9% of prompt changes are documented in commit messages, changes can introduce logical inconsistencies, and misalignment often occurs between prompt changes and LLM responses. These insights emphasize the need for specialized testing frameworks, automated validation tools, and improved documentation practices to enhance the reliability of LLM-integrated applications., Comment: Accepted at MSR 2025

In this work, we study flavor-violating scalars (flavons) in a range of large masses that have not been explored previously. We model the interactions with an effective field theory formulation where the flavon is heavier than the top quark. In addition, we assume that the flavon only couples to fermions of the Standard Model in a flavor-changing way. As the flavon couples strongly to top quarks, same-sign and opposite-sign top quark pair signals can be explored in the search for those particles. Using parametrized neural networks, we show that it is possible to probe flavons with masses in the 200-1600 GeV range through their interactions with a top quark plus up and charm quarks for effective couplings of order 10^-2 TeV^-1 at the 14 TeV High-Luminosity LHC., Comment: 15 pages, 6 figures

Publicly available datasets are subject to new versions, with each new version potentially reflecting changes to the data. These changes may involve adding or removing attributes, changing data types, and modifying values or their semantics. Integrating these datasets into a database poses a significant challenge: how to keep track of the evolving database schema while incorporating different versions of the data sources? This paper presents a statistical methodology to validate the integration of 12 years of open-access datasets from Brazil's School Census, with a new version of the datasets released annually by the Brazilian Ministry of Education (MEC). We employ various statistical tests to find matching attributes between datasets from a specific year and their potential equivalents in datasets from later years. The results show that by using the Kolmogorov-Smirnov test we can successfully match columns from different dataset versions in about 90% of cases., Comment: Accepted for presentation at the Simposio Brasileiro de Bancos de Dados (SBBD) 2024

Software engineers have adopted the Dynamic Software Product Lines (DSPL) engineering practices to develop Dynamically Adaptable Software (DAS). DAS is seen as a DSPL application and must cope with a large number of configurations of features, Non-functional Requirements (NFRs), and contexts. However, the accurate representation of the impact of features over NFRs and contexts for the identification of optimal configurations is not a trivial task. Software engineers need to have domain knowledge and design DAS before deploying to satisfy those requirements. Aiming to handle them, we proposed an approach of Trade-off Analysis for DSPL at design-time, named ToffA-DSPL. It deals with the configuration selection process considering interactions between NFRs and contexts. We performed an exploratory study based on simulations to identify the usefulness of the ToffA-DSPL approach. In general, the configurations suggested by ToffA-DSPL provide high satisfaction levels of NFRs. Based on simulations, we evidenced that our approach aims to explore reuse and is useful for generating valid and optimal configurations. In addition, ToffA-DSPL enables software engineers to conduct trade-off analysis, evaluate changes in the context feature, and define an adaptation model from optimal configurations found in the analysis.

Turnover consists of moving into and out of professional employees in the company in a given period. Such a phenomenon significantly impacts the software industry since it generates knowledge loss, delays in the schedule, and increased costs in the final project. Despite the efforts made by researchers and professionals to minimize the turnover, more studies are needed to understand the motivation that drives Software Engineers to leave their jobs and the main strategies CEOs adopt to retain these professionals in software development companies. In this paper, we contribute a mixed methods study involving semi-structured interviews with Software Engineers and CEOs to obtain a wider opinion of these professionals about turnover and a subsequent validation survey with additional software engineers to check and review the insights from interviews. In studying such aspects, we identified 19 different reasons for software engineers' turnover and 18 more efficient strategies used in the software development industry to reduce it. Our findings provide several implications for industry and academia, which can drive future research.

As Large Language Models (LLMs), including ChatGPT and analogous systems, continue to advance, their robust natural language processing capabilities and diverse applications have garnered considerable attention. Nonetheless, despite the increasing acknowledgment of the convergence of Artificial Intelligence (AI) and Software Engineering (SE), there is a lack of studies involving the impact of this convergence on the practices and perceptions of software developers. Understanding how software developers perceive and engage with AI tools, such as ChatGPT, is essential for elucidating the impact and potential challenges of incorporating AI-driven tools in the software development process. In this paper, we conducted a survey with 207 software developers to understand the impact of ChatGPT on software quality, productivity, and job satisfaction. Furthermore, the study delves into developers' expectations regarding future adaptations of ChatGPT, concerns about potential job displacement, and perspectives on regulatory interventions., Comment: 31 pages, 9 figures

The search for double Higgs production in $b\bar{b}W^+W^-$, where both $W$ bosons decay to leptons, has been rehabilitated as a good option to look for that key process to the Standard Model scalar sector study in the LHC. The missing neutrinos, however, hinder the reconstruction of useful information like the Higgs pair mass, which is very sensitive to the trilinear Higgs self-coupling. We present a solution to that problem using a Variational Autoencoder for Regression (VAER) to reconstruct the Higgs and top pairs decays $hh,t\bar{t}\to b\bar{b}W^+W^-\to b\bar{b}\ell^+\ell^{\prime -}\nu_\ell\bar{\nu}_{\ell^\prime}$. The algorithm predicts the invariant mass of non-resonant $hh$ irrespective of the trilinear coupling, even for events whose Higgs self-couplings were never presented to it. VAER is also able to identify a new Higgs resonance in an unsupervised way, showing generalization power for events not presented in its training phase. Finally, we demonstrate that VAER prediction is as useful to statistical inference as ground truth simulated distributions by computing a $\chi^2$ between trilinear coupling hypotheses based on binned invariant mass distributions of $b\bar{b}\ell^+\ell^{\prime -}\nu_\ell\bar{\nu}_{\ell^\prime}$., Comment: 29 pages, 10 figures, 2 tables

Software Quality Assurance (SQA) Engineers are responsible for assessing a product during every phase of the software development process to ensure that the outcomes of each phase and the final product possess the desired qualities. In general, a great SQA engineer needs to have a different set of abilities from development engineers to effectively oversee the entire product development process from beginning to end. Recent empirical studies identified important attributes of software engineers and managers, but the quality assurance role is overlooked. As software quality aspects have become more of a priority in the life cycle of software development, employers seek professionals that best suit the company's objectives and new graduates desire to make a valuable contribution through their job as an SQA engineer, but what makes them great? We addressed this knowledge gap by conducting 25 semi-structured interviews and 363 survey respondents with software quality assurance engineers from different companies around the world. We use the data collected from these activities to derive a comprehensive set of attributes that are considered important. As a result of the interviews, twenty-five attributes were identified and grouped into five main categories: personal, social, technical, management, and decision-making attributes. Through a rating survey, we confirmed that the distinguishing characteristics of great SQA engineers are curiosity, the ability to communicate effectively, and critical thinking skills. This work will guide further studies with SQA practitioners, by considering contextual factors and providing some implications for research and practice., Comment: 17 pages, 6 figures, 12 tables

The presence of bugs in video games can bring significant consequences for developers. To avoid these consequences, developers can leverage gameplay videos to identify and fix these bugs. Video hosting websites such as YouTube provide access to millions of game videos, including videos that depict bug occurrences, but the large amount of content can make finding bug instances challenging. We present an automated approach that uses machine learning to predict whether a segment of a gameplay video contains the depiction of a bug. We analyzed 4,412 segments of 198 gameplay videos to predict whether a segment contains an instance of a bug. Additionally, we investigated how our approach performs when applied across different specific genres of video games and on videos from the same game. We also analyzed the videos in the dataset to investigate what characteristics of the visual features might explain the classifier's prediction. Finally, we conducted a user study to examine the benefits of our automated approach against a manual analysis. Our findings indicate that our approach is effective at detecting segments of a video that contain bugs, achieving a high F1 score of 0.88, outperforming the current state-of-the-art technique for bug classification of gameplay video segments.

Traditionally, energy efficiency research has focused on reducing energy consumption at the hardware level and, more recently, in the design and coding phases of the software development life cycle. However, software testing's impact on energy consumption did not receive attention from the research community. Specifically, how test code design quality and test smell (e.g., sub-optimal design and bad practices in test code) impact energy consumption has not been investigated yet. This study examined 12 Apache projects to analyze the association between test smell and its effects on energy consumption in software testing. We conducted a mixed-method empirical analysis from two dimensions; software (data mining in Apache projects) and developers' views (a survey of 62 software practitioners). Our findings show that: 1) test smell is associated with energy consumption in software testing. Specifically smelly part of a test case consumes 10.92\% more energy compared to the non-smelly part. 2) certain test smells are more energy-hungry than others, 3) refactored test cases tend to consume less energy than their smelly counterparts, and 4) most developers lack knowledge about test smells' impact on energy consumption. We conclude the paper with several observations that can direct future research and developments.

Mental health disorders affect a large number of people, leading to many lives being lost every year. These disorders affect struggling individuals and businesses whose productivity decreases due to days of lost work or lower employee performance. Recent studies provide alarming numbers of individuals who suffer from mental health disorders, e.g., depression and anxiety, in particular contexts, such as academia. In the context of the software industry, there are limited studies that aim to understand the presence of mental health disorders and the characteristics of jobs in this context that can be triggers for the deterioration of the mental health of software professionals. In this paper, we present the results of a survey with 500 software professionals. We investigate different aspects of their mental health and the characteristics of their work to identify possible triggers of mental health deterioration. Our results provide the first evidence that mental health is a critical issue to be addressed in the software industry, as well as raise the direction of changes that can be done in this context to improve the mental health of software professionals., Comment: 12 pages, 3 figures

For companies producing related products, a Software Product Line (SPL) is a software reuse method that improves time-to-market and software quality, achieving substantial cost reductions.These benefits do not come for free. It often takes years to re-architect and re-engineer a codebase to support SPL and, once adopted, it must be maintained. Current SPL practice relies on a collection of tools, tailored for different reengineering phases, whose output developers must coordinate and integrate. We present Foundry, a general automated approach for leveraging software transplantation to speed conversion to and maintenance of SPL. Foundry facilitates feature extraction and migration. It can efficiently, repeatedly, transplant a sequence of features, implemented in multiple files. We used Foundry to create two valid product lines that integrate features from three real-world systems in an automated way. Moreover, we conducted an experiment comparing Foundry's feature migration with manual effort. We show that Foundry automatically migrated features across codebases 4.8 times faster, on average, than the average time a group of SPL experts took to accomplish the task.

Software testing is a mandatory activity in any serious software development process, as bugs are a reality in software development. This raises the question of quality: good tests are effective in finding bugs, but until a test case actually finds a bug, its effectiveness remains unknown. Therefore, determining what constitutes a good or bad test is necessary. This is not a simple task, and there are a number of studies that identify different characteristics of a good test case. A previous study evaluated 29 hypotheses regarding what constitutes a good test case, but the findings are based on developers' beliefs, which are subjective and biased. In this paper we investigate eight of these hypotheses, through an extensive empirical study based on open software repositories. Despite our best efforts, we were unable to find evidence that supports these beliefs. This indicates that, although these hypotheses represent good software engineering advice, they do not necessarily mean that they are enough to provide the desired outcome of good testing code., Comment: 12 pages, 1 figure, 3 tables

We explore single and double flavor-violating scalar (flavon) production at the 13 and 14 TeV LHC in an effective field theory formulation where flavons always change the flavor of the Standard Model fermions. When those scalars couple to mass, their flavor-changing couplings to top quarks and tau leptons are favored. Focusing on the mass region below the top-quark mass, we find couplings that fit the muon $(g-2)$ discrepancy and avoid several current experimental constraints. We determine the potential of the LHC to exclude or discover such a new physics scenario with clean signatures consisting of same-sign tau leptons and the simultaneous observation of resonances in the tau plus electron or muon invariant mass. We found that in the double production mode, effective couplings down to order $10^{-2}$ TeV$^{-1}$ can be probed for flavon masses in the 10--170 GeV range at the 14 TeV HL-LHC, but couplings down to 0.1 TeV$^{-1}$ can already be excluded at 95\% confidence level with data collected from the 13 TeV LHC in the same mass interval. We also explore the impact of sizeable diagonal flavon couplings on the prospects of LHC for the signals we propose., Comment: 33 pages, 13 figures, 2 tables. Accepted for publication on PRD

Unmanned Aerial Vehicles (UAVs) offer promising potential as communications node carriers, providing on-demand wireless connectivity to users. While existing literature presents various wireless channel models, it often overlooks the impact of UAV heading. This paper provides an experimental characterization of the Air-to-Ground (A2G) and Ground-to-Air (G2A) wireless channels in an open environment with no obstacles nor interference, considering the distance and the UAV heading. We analyze the received signal strength indicator and the TCP throughput between a ground user and a UAV, covering distances between 50~m and 500~m, and considering different UAV headings. Additionally, we characterize the antenna's radiation pattern based on UAV headings. The paper provides valuable perspectives on the capabilities of UAVs in offering on-demand and dynamic wireless connectivity, as well as highlights the significance of considering UAV heading and antenna configurations in real-world scenarios.

Digital twins have been emerging as a hybrid approach that combines the benefits of simulators with the realism of experimental testbeds. The accurate and repeatable set-ups replicating the dynamic conditions of physical environments, enable digital twins of wireless networks to be used to evaluate the performance of next-generation networks. In this paper, we propose the Position-based Machine Learning Propagation Loss Model (P-MLPL), enabling the creation of fast and more precise digital twins of wireless networks in ns-3. Based on network traces collected in an experimental testbed, the P-MLPL model estimates the propagation loss suffered by packets exchanged between a transmitter and a receiver, considering the absolute node's positions and the traffic direction. The P-MLPL model is validated with a test suite. The results show that the P-MLPL model can predict the propagation loss with a median error of 2.5 dB, which corresponds to 0.5x the error of existing models in ns-3. Moreover, ns-3 simulations with the P-MLPL model estimated the throughput with an error up to 2.5 Mbit/s, when compared to the real values measured in the testbed.

We study the lepton-induced resonant production of color-adjoint leptons (leptogluons) at the LHC employing the lepton parton density function of the proton. We demonstrate that this production mechanism can be useful to extend the LHC ability to search for leptogluons beyond purely quark/gluon initiated production processes up to ~ 3.5 TeV leptogluon masses and O(1) TeV compositeness scales. Discerning leptogluons from scalar and vector leptoquarks is also possible in this channel, given a data sample containing the order of 100 signal events. We argue that the resonant channel can be combined with leptogluon pair and associated leptogluon-lepton productions to boost exclusion limits and discovery prospects at the LHC., Comment: 15 pages, 8 figures

The whiteboard plays a crucial role in the day-to-day lives of software architects, as they frequently will organize meetings at the whiteboard to discuss a new architecture, some proposed changes to the architecture, a mismatch between the architecture and the code, and more. While much has been studied about software architects, the architectures they produce, and how they produce them, a detailed understanding of these whiteboards meetings is still lacking. In this paper, we contribute a mixed-methods study involving semi-structured interviews and a subsequent survey to understand the perceptions of software architects on whiteboard architecture meetings. We focus on five aspects: (1) why do they hold these meetings, what is the impact of the experience levels of the participants in these meetings, how do the architects document the meetings, what kinds of changes are made after the meetings have concluded and their results are moved to implementation, and what role do digital whiteboards plays? In studying these aspects, we identify 12 observations related to both technical aspects and social aspects of the meetings. These insights have implications for further research, offer concrete advice to practitioners, provide guidance for future tool design, and suggest ways of educating future software architects.

Computational notebooks, such as Jupyter, have been widely adopted by data scientists to write code for analyzing and visualizing data. Despite their growing adoption and popularity, there has been no thorough study to understand Jupyter development challenges from the practitioners' point of view. This paper presents a systematic study of bugs and challenges that Jupyter practitioners face through a large-scale empirical investigation. We mined 14,740 commits from 105 GitHub open-source projects with Jupyter notebook code. Next, we analyzed 30,416 Stack Overflow posts which gave us insights into bugs that practitioners face when developing Jupyter notebook projects. Finally, we conducted nineteen interviews with data scientists to uncover more details about Jupyter bugs and to gain insights into Jupyter developers' challenges. We propose a bug taxonomy for Jupyter projects based on our results. We also highlight bug categories, their root causes, and the challenges that Jupyter practitioners face.

The creation of digital twins of experimental testbeds allows the validation of novel wireless networking solutions and the evaluation of their performance in realistic conditions, without the cost, complexity and limited availability of experimental testbeds. Current trace-based simulation approaches for ns-3 enable the repetition and reproduction of the same exact conditions observed in past experiments. However, they are limited by the fact that the simulation setup must exactly match the original experimental setup, including the network topology, the mobility patterns and the number of network nodes. In this paper, we propose the Machine Learning based Propagation Loss (MLPL) module for ns-3. Based on network traces collected in an experimental testbed, the MLPL module estimates the propagation loss as the sum of a deterministic path loss and a stochastic fast-fading loss. The MLPL module is validated with unit tests. Moreover, we test the MLPL module with real network traces, and compare the results obtained with existing propagation loss models in ns-3 and real experimental results. The results obtained show that the MLPL module can accurately predict the propagation loss observed in a real environment and reproduce the experimental conditions of a given testbed, enabling the creation of digital twins of wireless network environments in ns-3.

We analyze the impact of the recently released CDF-II measurement of $W$ mass on the SMEFT analyses of the electroweak precision data as well as Higgs and electroweak diboson productions. We work in the Hagiwara, Ishihara, Szalapski, and Zeppenfeld basis in which eight generation-independent operators enter in the electroweak precision data at tree level and, unlike in the Warsaw basis, the analysis of that set of data constrains all the eight Wilson coefficients, without the need of combination with Higgs or electroweak diboson data results. We show that in the global analysis the determination of the coefficients of all operators which do not enter the electroweak precision data are barely affected by the new $M_W$ determination., Comment: 6 pages, 3 figures

Unmanned Aerial Vehicles (UAVs) acting as Flying Access Points (FAPs) are being used to provide on-demand wireless connectivity in extreme scenarios. Despite ongoing research, the optimization of UAVs' positions according to dynamic users' traffic demands remains challenging. We propose the Traffic-aware UAV Placement Algorithm (TUPA), which positions a UAV acting as FAP according to the users' traffic demands, in order to maximize the network utility. Using a DRL approach enables the FAP to autonomously learn and adapt to dynamic conditions and requirements of networking scenarios. Moreover, the proposed DRL methodology allows TUPA to generalize knowledge acquired during training to unknown combinations of users' positions and traffic demands, with no additional training. TUPA is trained and evaluated using network simulator ns-3 and ns3-gym framework. The results demonstrate that TUPA increases the network utility, compared to baseline solutions, increasing the average network utility up to 4x in scenarios with heterogeneous traffic demands.

As Deep learning (DL) systems continuously evolve and grow, assuring their quality becomes an important yet challenging task. Compared to non-DL systems, DL systems have more complex team compositions and heavier data dependency. These inherent characteristics would potentially cause DL systems to be more vulnerable to bugs and, in the long run, to maintenance issues. Code smells are empirically tested as efficient indicators of non-DL systems. Therefore, we took a step forward into identifying code smells, and understanding their impact on maintenance in this comprehensive study. This is the first study on investigating code smells in the context of DL software systems, which helps researchers and practitioners to get a first look at what kind of maintenance modification made and what code smells developers have been dealing with. Our paper has three major contributions. First, we comprehensively investigated the maintenance modifications that have been made by DL developers via studying the evolution of DL systems, and we identified nine frequently occurred maintenance-related modification categories in DL systems. Second, we summarized five code smells in DL systems. Third, we validated the prevalence, and the impact of our newly identified code smells through a mixture of qualitative and quantitative analysis. We found that our newly identified code smells are prevalent and impactful on the maintenance of DL systems from the developer's perspective.

The increasing complexity of recent Wi-Fi amendments is making optimal Rate Adaptation (RA) a challenge. The use of classic algorithms or heuristic models to address RA is becoming unfeasible due to the large combination of configuration parameters along with the variability of the wireless channel. Machine Learning-based solutions have been proposed in the state of art, to deal with this complexity. However, they typically use complex models and their implementation in real scenarios is difficult. We propose a simple Deep Reinforcement Learning approach for the automatic RA in Wi-Fi networks, named Data-driven Algorithm for Rate Adaptation (DARA). DARA is standard-compliant. It dynamically adjusts the Wi-Fi Modulation and Coding Scheme (MCS) solely based on the observation of the Signal-to-Noise Ratio (SNR) of the received frames at the transmitter. Our simulation results show that DARA achieves up to 15\% higher throughput when compared with Minstrel High Throughput (HT) and equals the performance of the Ideal Wi-Fi RA algorithm.

We present a comprehensive study of the electroweak interactions using the available Higgs and electroweak diboson production results from LHC Runs 1 and 2 as well as the electroweak precision data, in terms of the dimension-six operators. Under the assumption that no new tree level sources of flavor violation nor violation of universality of the weak current is introduced, the analysis involves 21 operators. We assess the impact of the data on kinematic distributions for the Higgs production at the LHC by comparing the results obtained including the simplified template cross section data with those in which only total Higgs signal strengths are considered. We also compare the results obtained when including the dimension-six anomalous contributions to order $1/\Lambda^2$ and to order $1/\Lambda^4$. As an illustration of the LHC potential to indirectly learn about specific forms of new physics, we adapt the analysis to constrain the parameter space for a few simple extensions of the standard model which generate a subset of the dimension-six operators at tree level., Comment: version to appear in PRD with updates and corrections

Bugs that persist into releases of video games can have negative impacts on both developers and users, but particular aspects of testing in game development can lead to difficulties in effectively catching these missed bugs. It has become common practice for developers to apply updates to games in order to fix missed bugs. These updates are often accompanied by notes that describe the changes to the game included in the update. However, some bugs reappear even after an update attempts to fix them. In this paper, we develop a taxonomy for bug types in games that is based on prior work. We examine 12,122 bug fixes from 723 updates for 30 popular games on the Steam platform. We label the bug fixes included in these updates to identify the frequency of these different bug types, the rate at which bug types recur over multiple updates, and which bug types are treated as more severe. Additionally, we survey game developers regarding their experience with different bug types and what aspects of game development they most strongly associate with bug appearance. We find that Information bugs appear the most frequently in updates, while Crash bugs recur the most frequently and are often treated as more severe than other bug types. Finally, we find that challenges in testing, code quality, and bug reproduction have a close association with bug persistence. These findings should help developers identify which aspects of game development could benefit from greater attention in order to prevent bugs. Researchers can use our results in devising tools and methods to better identify and address certain bug types.