arXiv subjects

Sven Apel

Publications and source records attributed to Sven Apel.

At least 19 recordsLinked to original sources

arXiv2026

Developer Attitudes and Practices Towards Optimizing Software Energy Consumption

Context: Software significantly influences the efficiency with which hardware resources are utilized, yet software energy consumption is seldom treated as a first-class concern in day-to-day development practice. Objective: This study investigates professional developers' attitudes, decision-making, and development practices related to software energy consumption, with particular emphasis on how energy considerations are recognized, assessed, and acted upon during software development. Method: To this end, we conduct an online survey with 134 software developers. Our study combines quantitative analyses with a qualitative open-card sorting of free-text responses to characterize perceptions, practices, and reasoning patterns around energy consumption. Findings: Energy consumption is explicitly considered in only a minority of projects. More commonly, developers influence energy use indirectly by optimizing proxy properties such as execution time and CPU utilization. Responses to scenario-based questions reveal systematic blind spots in this mental model, including cases in which performance improvements increase energy consumption or exhibit no correlation. We also identify organizational disincentives, limited tooling, and educational gaps as major barriers to adoption. Implications: (1) Institutionalize energy-aware approaches through visible flagship deployments that demonstrate value, (2) expand research and education on energy-performance trade-offs, and (3) develop practical, developer-oriented measurement and feedback tools that lower adoption barriers.

cs.SE↗

arXiv2026

Motivations and Barriers to Communicating Software Engineering Research: Insights from Early Career Researchers

Science communication is increasingly becoming a part of modern research careers, involving researchers to disseminate knowledge, engage broader communities, and increase the societal impact of their work. Despite its growing importance, little is known about how early-career software engineering researchers perceive and navigate science communication in practice. In this paper, we investigate how PhD students in software engineering experience science communication. We conducted semi-structured interviews with 18 doctoral candidates from diverse international backgrounds. Using thematic analysis, we examine three interconnected dimensions: motivations, communication channels, and barriers. Our findings reveal a strong tension between aspiration and practice. Participants were highly motivated to engage in science communication due to opportunities for collaboration, professional recognition, broader impact, and advocacy for themselves and their research. However, translating these motivations into action was frequently constrained by social anxiety, uncertainty regarding appropriate audiences and communication venues, limited feedback mechanisms, insufficient institutional guidance, and challenges associated with navigating an increasingly fragmented communication landscape. Our findings highlight the need for tailored, balanced support systems that empower software engineering PhD students to engage in science communication effectively and confidently across diverse cultural and institutional environments. We outline practical implications that offer initial guidance for addressing these challenges in future work.

cs.SE↗

arXiv2026

Predicting Program Comprehension with Foundation Models of Human Cognition

Software engineering depends on the ability of developers to understand code, yet predicting how they do so remains an open challenge despite decades of research. Existing approaches rely either on simplified proxy measures that limit accuracy or on non-trivial measurements requiring elaborate experimental setups that are difficult to scale and apply in practice. In contrast, recent work in psychology suggests an alternative perspective: Instead of modeling task-specific phenomena directly, human behavior can be captured through cognitive regularities learned from large-scale behavioral data. This idea treats complex human behavior as the observable outcome of underlying cognitive processes that manifest consistently across tasks and domains. In this paper, we explore this perspective in the context of program comprehension. We evaluate Centaur, a foundation model trained on 160 general psychological experiments, on 9 previously published program-comprehension studies. We assess how well its predicted response distributions align with human response data and compare Centaur's performance to its base model, Llama 3.1. To better understand the source of its performance, we conduct ablation studies to isolate the contribution of different sources of information, such as the code artifacts, task-related context, and prior trials and participant responses. In a nutshell, we find that Centaur more closely aligns with human response patterns than its base model, is significantly less reliant on information from prior trials and responses, and benefits more from task-related information. These findings suggest that behavioral patterns learned from general psychological data can transfer to complex software engineering tasks such as program comprehension. More broadly, they point toward foundation models of human cognition as a basis for modeling developer behavior in software engineering.

cs.SE↗

arXiv2026

A Mechanistic Lens on Semantic Conflicts: Using Activation Patching to Understand LLM Behavior

Large language models (LLMs) are increasingly used in software-engineering tasks processing executable code and non-executable semantic cues such as comments or identifiers. These two sources can conflict when semantic cues suggest different program behavior than the code itself. It remains unclear how such semantic conflicts affect LLM behavior and which source dominates their outputs. We present the first controlled, mechanistic study of LLM behavior under semantic conflicts. To this end, we construct 45 Python snippet triplets that isolate conflicts by varying either semantic cues or implementation while keeping token-aligned pairs for causal intervention. We evaluate four open-weight LLMs on two tasks (output prediction and unit-test generation) using behavioral performance measures and residual-stream activation patching to identify token-layer states that causally contribute to behavioral differences between aligned and conflicting inputs. Our results show that semantic conflicts significantly reduce execution-grounded correctness in both tasks and that all tested LLMs often follow misleading semantic cues. Residual-stream activation patching reveals a consistent pattern for final-output prediction: The changed cue/code region and a small set of intermediate tokens carry most of the recoverable causal signal before aggregation near the output readout. For unit-test generation, this pattern extends beyond the prompt, showing that conflict-related information is recoverable at generated sites before producing expected values. Overall, our findings show that semantic conflicts affect program comprehension and downstream tasks, with relevant information concentrated in a small number of causally active residual-stream states, and demonstrate a framework for mechanistically analyzing how LLMs integrate code-related information under controlled semantic variations.

cs.SE↗

arXiv2026

Neural Signatures of Programming Expertise: Classifying Programmer Skill Levels Using EEG Data

Accurately assessing a programmer's skill level is critical for hiring, team composition, and performance evaluation in the software industry. Conventional methods, such as coding tests or interviews, often fail to capture the full spectrum of cognitive abilities underlying programming expertise. This study explores using electroencephalography (EEG) and machine learning to investigate neural correlates of programming skill. We analyzed an existing EEG dataset recorded during code comprehension from 37 programmers with 1 to 30 years of experience (8.1 +/- 6.3 years) to examine relationships between neural activity and expertise. Additionally, we conducted classification experiments using Random Forest classifiers with diverse features for binary (experts vs. novices) and multi-class (experts, intermediates, novices) setups. We identified EEG features and brain regions associated with programming expertise. Specifically, EEG entropy showed the strongest correlation with skill level. Furthermore, experts' brains were characterized by highly localized centro-frontal activation, whereas frontal activation in other groups was part of a more distributed network. Regarding classification, our setup achieved an average accuracy of 91.83% (binary) and 78.15% (multi-class) in stratified 10-fold cross-validation, while leave-one-subject-out validation achieved 85.00% and 58.80%, respectively. Individual frequency bands outperformed full-spectrum analyses, and both program comprehension and resting-state data yielded strong results. These findings demonstrate that EEG features effectively capture neural correlates across different skill levels and highlight the potential of neural data to complement traditional methods of skill assessment.

cs.HC↗

arXiv2026

Software Engineering Podcasts: An Empirical Study of Their Potential as a Research Resource

Podcasts have become an increasingly popular medium for knowledge sharing within the software engineering (SE) community, offering insights into industry developments and the perspectives of professionals with different backgrounds. As this medium grows, it presents a potentially valuable resource not only for practitioners but also for researchers seeking to understand the evolving field. However, little is known about the actual content of SE podcasts or how they are perceived and used by researchers. This study systematically explores the SE podcast landscape, analyzing its content and surveying researchers to assess how podcasts can serve as a meaningful resource for advancing empirical software engineering research.

cs.SE↗

arXiv2026

Harnessing Hype to Teach Empirical Thinking: An Experience With AI Coding Assistants

Software engineering students often struggle to appreciate empirical methods and hypothesis-driven inquiry, especially when taught in theoretical terms. This experience report explores whether grounding empirical learning in hype-driven technologies can make these concepts more accessible and engaging. We conducted a one-semester seminar framed around the currently popular topic of AI coding assistants, which attracted unusually high student interest. The course combined hands-on sessions using AI coding assistants with small, student-designed empirical studies. Classroom observations and survey responses suggest that the hype topic sparked curiosity and critical thinking. Students engaged with the AI coding assistants while questioning their limitations -- developing the kind of empirical thinking needed to assess claims about emerging technologies. Key lessons: (1) Hype-driven topics can lower barriers to abstract concepts like empirical research; (2) authentic hands-on development tasks combined with ownership of inquiry foster critical engagement; and (3) a single seminar can effectively teach both technical and research skills.

cs.SE↗

arXiv2026

From Restructuring to Stabilization: A Large-Scale Experiment on Iterative Code Readability Refactoring with Large Language Models

Large language models (LLMs) are increasingly used for automated code refactoring tasks. Although these models can quickly refactor code, the quality may exhibit inconsistencies and unpredictable behavior. In this article, we systematically study the capabilities of LLMs for code refactoring with a specific focus on improving code readability. We conducted a large-scale experiment using GPT5.1 with 230 Java snippets, each systematically varied and refactored regarding code readability across five iterations under three different prompting strategies. We categorized fine-grained code changes during the refactoring into implementation, syntactic, and comment-level transformations. Subsequently, we investigated the functional correctness and tested the robustness of the results with novel snippets. Our results reveal three main insights: First, iterative code refactoring exhibits an initial phase of restructuring followed by stabilization. This convergence tendency suggests that LLMs possess an internalized understanding of an "optimally readable" version of code. Second, convergence patterns are fairly robust across different code variants. Third, explicit prompting toward specific readability factors slightly influences the refactoring dynamics. These insights provide an empirical foundation for assessing the reliability of LLM-assisted code refactoring, which opens pathways for future research, including comparative analyses across models and a systematic evaluation of additional software quality dimensions in LLM-refactored code.

cs.SE↗

arXiv2026

Multi-Location Software Model Completion

In model-driven engineering and beyond, software models are key development artifacts. In practice, they often grow to substantial size and complexity, undergoing thousands of modifications over time due to evolution, refactoring, and maintenance. The rise of AI has sparked interest in how software modeling activities can be automated. Recently, LLM-based approaches for software model completion have been proposed, however, the state of the art supports only single-location model completion by predicting changes at a specific location. Going beyond, we aim to bridge the gap toward handling coordinated changes that span multiple locations across large, complex models. Specifically, we propose a novel global embedding-based next focus predictor, NextFocus, which is capable of multi-location model completion for the first time. The predictor consists of a neural network with an attention mechanism that is trained on historical software model evolution data. Starting from an existing change, it predicts further model elements to change, potentially spanning multiple parts of the model. We evaluate our approach on multi-location model changes that have actually been performed by developers in real-world projects. NextFocus achieves promising results for multi-location model completion, even when changes are heavily spread across the model. It achieves an average Precision@k score of 0.98 for $k \leq 10$, significantly outperforming the three baseline approaches.

cs.SE↗

arXiv2025

Detecting Performance-Relevant Changes in Configurable Software Systems

Performance is a volatile property of a software system and frequent performance profiling is required to keep the knowledge about a software system's performance behavior up to date. Repeating all performance measurements after every revision is a cost-intensive task, especially in the presence of configurability, where one has to measure multiple configurations to obtain a comprehensive picture. Configuration sampling is a common approach to control the measurement cost. However, it cannot guarantee completeness and might miss performance regressions, especially if they only affect few configurations. As an alternative to solve the cost reduction problem, we present ConfFLARE: ConfFLARE estimates whether a change potentially impacts performance by identifying data-flow interactions with performance-relevant code and extracts which software features participate in such interactions. Based on these features, we can select a subset of relevant configurations to focus performance profiling efforts on. In a study conducted on both, synthetic and real-world software systems, ConfFLARE correctly detects performance regressions in almost all cases and identifies relevant features in all but two cases, reducing the number of configurations to be tested on average by $79\%$ for synthetic and by $70\%$ for real-world regression scenarios saving hours of performance testing time.

cs.SE↗

arXiv2025

How do Humans and LLMs Process Confusing Code?

Already today, humans and programming assistants based on large language models (LLMs) collaborate in everyday programming tasks. Clearly, a misalignment between how LLMs and programmers comprehend code can lead to misunderstandings, inefficiencies, low code quality, and bugs. A key question in this space is whether humans and LLMs are confused by the same kind of code. This would not only guide our choices of integrating LLMs in software engineering workflows, but also inform about possible improvements of LLMs. To this end, we conducted an empirical study comparing an LLM to human programmers comprehending clean and confusing code. We operationalized comprehension for the LLM by using LLM perplexity, and for human programmers using neurophysiological responses (in particular, EEG-based fixation-related potentials). We found that LLM perplexity spikes correlate both in terms of location and amplitude with human neurophysiological responses that indicate confusion. This result suggests that LLMs and humans are similarly confused about the code. Based on these findings, we devised a data-driven, LLM-based approach to identify regions of confusion in code that elicit confusion in human programmers.

cs.SE↗

arXiv2025

Automata Learning -- Expect Delays!

This paper studies active automata learning (AAL) in the presence of stochastic delays. We consider Mealy machines that have stochastic delays associated with each transition and explore how the learner can efficiently arrive at faithful estimates of those machines, the precision of which crucially relies on repetitive sampling of transition delays. While it is possible to na\"ively integrate the delay sampling into AAL algorithms such as $L^*$, this leads to considerable oversampling near the root of the state space. We address this problem by separating conceptually the learning of behavior and delays such that the learner uses the information gained while learning the logical behavior to arrive at efficient input sequences for collecting the needed delay samples. We put emphasis on treating cases in which identical input/output behaviors might stem from distinct delay characteristics. Finally, we provide empirical evidence that our method outperforms the na\"ive baseline across a wide range of benchmarks and investigate its applicability in a realistic setting by studying the join order in a relational database.

cs.FL↗

arXiv2025

From Developer Pairs to AI Copilots: A Comparative Study on Knowledge Transfer

Knowledge transfer is fundamental to human collaboration and is therefore common in software engineering. Pair programming is a prominent instance. With the rise of AI coding assistants, developers now not only work with human partners but also, as some claim, with AI pair programmers. Although studies confirm knowledge transfer during human pair programming, its effectiveness with AI coding assistants remains uncertain. To analyze knowledge transfer in both human-human and human-AI settings, we conducted an empirical study where developer pairs solved a programming task without AI support, while a separate group of individual developers completed the same task using the AI coding assistant GitHub Copilot. We extended an existing knowledge transfer framework and employed a semi-automated evaluation pipeline to assess differences in knowledge transfer episodes across both settings. We found a similar frequency of successful knowledge transfer episodes and overlapping topical categories across both settings. Two of our key findings are that developers tend to accept GitHub Copilot's suggestions with less scrutiny than those from human pair programming partners, but also that GitHub Copilot can subtly remind developers of important code details they might otherwise overlook.

cs.SE↗

arXiv2025

The Silent Scientist: When Software Research Fails to Reach Its Audience

If software research were a performance, it would be a thoughtful theater play -- full of rich content but confined to the traditional stage of academic publishing. Meanwhile, its potential audience is immersed in engaging on-demand experiences, leaving the theater half-empty, and the research findings lost in the wings. As long as this remains the case, discussions about research relevance and impact lack meaningful context.

cs.SE↗

arXiv2025

Pragmatic Reasoning improves LLM Code Generation

Pragmatic reasoning helps interlocutors infer intended meaning from ambiguous or underspecified messages by considering shared context and counterfactual alternatives. Similar challenges arise in natural language-to-code generation, where user instructions often admit multiple plausible candidate programs. However, direct RSA-style inference is difficult because it requires probability estimation over large spaces of programs and alternative instructions. We propose CodeRSA, an RSA-motivated reranking method that makes pragmatic reasoning tractable through local pragmatic contests among sampled code candidates. CodeRSA constructs candidate-induced alternative instructions and estimates which candidates are most distinctively supported by the original instruction, avoiding global normalization over the full program-instruction space. We evaluate CodeRSA on HumanEval+, MBPP+, and BigCodeBench using four open-weight instruction-following models. CodeRSA achieves the strongest average accuracy in 10 of 12 model-benchmark settings and remains competitive in the remaining cases. Further analyses show that its gains come from combining local pairwise pragmatic comparison with broader global support, suggesting a scalable direction for language-to-code reranking under natural-language uncertainty.

cs.CL↗

arXiv2024

Hints Help Finding and Fixing Bugs Differently in Python and Text-based Program Representations

With the recent advances in AI programming assistants such as GitHub Copilot, programming is not limited to classical programming languages anymore--programming tasks can also be expressed and solved by end-users in natural text. Despite the availability of this new programming modality, users still face difficulties with algorithmic understanding and program debugging. One promising approach to support end-users is to provide hints to help them find and fix bugs while forming and improving their programming capabilities. While it is plausible that hints can help, it is unclear which type of hint is helpful and how this depends on program representations (classic source code or a textual representation) and the user's capability of understanding the algorithmic task. To understand the role of hints in this space, we conduct a large-scale crowd-sourced study involving 753 participants investigating the effect of three types of hints (test cases, conceptual, and detailed), across two program representations (Python and text-based), and two groups of users (with clear understanding or confusion about the algorithmic task). We find that the program representation (Python vs. text) has a significant influence on the users' accuracy at finding and fixing bugs. Surprisingly, users are more accurate at finding and fixing bugs when they see the program in natural text. Hints are generally helpful in improving accuracy, but different hints help differently depending on the program representation and the user's understanding of the algorithmic task. These findings have implications for designing next-generation programming tools that provide personalized support to users, for example, by adapting the programming modality and providing hints with respect to the user's skill level and understanding.

cs.SE↗

arXiv2024

Fixation-related potentials reveal that confusing program code elicits a late frontal positivity

As software pervades more and more areas of our professional and personal lives, there is an ever-increasing need to maintain software and for programmers to efficiently write and understand program code. In the first study of its kind, we analyze fixation-related potentials (FRPs) to explore the online processing of program code patterns that are confusing to programmers, but not to the computer (so-called atoms of confusion), and their underlying neurocognitive mechanisms in an ecologically valid setting. Relative to clean counterparts in program code without an atom of confusion, confusing code elicits a late frontal positivity of about 400 to 700 ms after first looking at the atom of confusion. This frontal positivity resembles an event-related potential (ERP) component found during natural language processing that is elicited by unexpected but plausible words in sentence context. Thus, we suggest that the brain engages similar neurocognitive mechanisms in response to unexpected and informative inputs in program code and in natural language. In both domains, these inputs update a comprehender's situation model, which is essential for information extraction from a quickly unfolding input. Our results have far-reaching implications for programming and pave the way for interdisciplinary collaborations between software engineering and psycholinguistics.

cs.SE↗

arXiv2024

Software Model Evolution with Large Language Models: Experiments on Simulated, Public, and Industrial Datasets

Modeling structure and behavior of software systems plays a crucial role in the industrial practice of software engineering. As with other software engineering artifacts, software models are subject to evolution. Supporting modelers in evolving software models with recommendations for model completions is still an open problem, though. In this paper, we explore the potential of large language models for this task. In particular, we propose an approach, RAMC, leveraging large language models, model histories, and retrieval-augmented generation for model completion. Through experiments on three datasets, including an industrial application, one public open-source community dataset, and one controlled collection of simulated model repositories, we evaluate the potential of large language models for model completion with RAMC. We found that large language models are indeed a promising technology for supporting software model evolution (62.30% semantically correct completions on real-world industrial data and up to 86.19% type-correct completions). The general inference capabilities of large language models are particularly useful when dealing with concepts for which there are few, noisy, or no examples at all.

cs.SE↗