arXiv subjects

Sherlock Licorish

Publications and source records attributed to Sherlock Licorish.

3 recordsLinked to original sources

arXiv2026

Teaching Empathy in Software Engineering Education in the Age of Artificial Intelligence

Empathy has been discussed as a relevant human capability in software engineering, particularly in activities that require understanding users, stakeholders, and the societal implications of technological systems. This relevance becomes more pronounced in the context of artificial intelligence, where software increasingly participates in decisions that affect diverse individuals and communities. However, limited guidance exists on how empathy can be integrated into technical software engineering education in ways that connect with the development of AI-enabled systems. This study investigates teaching practices that educators use to incorporate empathy into software engineering courses. Using qualitative analysis of educator-reported practices, we identified five categories through which empathy is operationalized within technical coursework: societal framing of AI systems, fairness and accessibility considerations in design and evaluation, representation of diverse users, stakeholder role awareness and responsibility, and structured reflection and feedback during development processes. The findings indicate that empathy can be embedded within core development activities rather than taught as a separate topic, enabling students to reason about bias, accessibility, accountability, and the societal consequences of AI technologies. These results contribute a structured view of how empathy-oriented practices can be incorporated into software engineering education to support the preparation of students who will develop AI-enabled systems.

cs.CY↗

arXiv2026

How Fair is Software Fairness Testing?

Software fairness testing is a central method for evaluating AI systems, yet the meaning of fairness is often treated as fixed and universally applicable. This vision paper positions fairness testing as culturally situated and examines the problem across three dimensions. First, fairness metrics encode particular cultural values while marginalizing others. Second, test datasets are predominantly designed from Western contexts, excluding knowledge systems grounded in oral traditions, Indigenous languages, and non-digital communities. Third, fairness testing raises ethical concerns, including the reliance on low-paid data labeling in the Global South, and associated with this, the environmental costs of training and deploying large-scale models, which disproportionately affect climate-vulnerable populations. Addressing these issues requires rethinking fairness testing beyond universal metrics and moving toward evaluation frameworks that respect cultural plurality and acknowledge the right to refuse algorithmic mediation.

cs.SE↗

arXiv2024

Optimizing Large Language Model Hyperparameters for Code Generation

Large Language Models (LLMs), such as GPT models, are increasingly used in software engineering for various tasks, such as code generation, requirements management, and debugging. While automating these tasks has garnered significant attention, a systematic study on the impact of varying hyperparameters on code generation outcomes remains unexplored. This study aims to assess LLMs' code generation performance by exhaustively exploring the impact of various hyperparameters. Hyperparameters for LLMs are adjustable settings that affect the model's behaviour and performance. Specifically, we investigated how changes to the hyperparameters: temperature, top probability (top_p), frequency penalty, and presence penalty affect code generation outcomes. We systematically adjusted all hyperparameters together, exploring every possible combination by making small increments to each hyperparameter at a time. This exhaustive approach was applied to 13 Python code generation tasks, yielding one of four outcomes for each hyperparameter combination: no output from the LLM, non executable code, code that fails unit tests, or correct and functional code. We analysed these outcomes for a total of 14,742 generated Python code segments, focusing on correctness, to determine how the hyperparameters influence the LLM to arrive at each outcome. Using correlation coefficient and regression tree analyses, we ascertained which hyperparameters influence which aspect of the LLM. Our results indicate that optimal performance is achieved with a temperature below 0.5, top probability below 0.75, frequency penalty above -1 and below 1.5, and presence penalty above -1. We make our dataset and results available to facilitate replication.

cs.SE↗