http://hdl.handle.net/1893/478 Electronic copies of Computing Science and Mathematics conference papers and proceedings. 2026-10-08T19:16:45Z http://hdl.handle.net/1893/38384 Title: What Do AI Agents Actually Change? An Empirical Taxonomy of Mutation Patterns in Performance-Improving Pull Requests Anonymous Anonymous Institution Author(s): Dovhoshliubnyi, llia; Soroush, Nima; Sami, Ashkan; Brownlee, Alexander Abstract: AI coding agents are black boxes: we cannot inspect how they generate code, but we can inspect what they change. This distinction matters for search-based software engineering (SBSE), where techniques such as genetic improvement depend on mutation operators that reflect how code is actually transformed. Of the 33,596 agent PRs in the AIDev dataset, less than 400 target performance (fewer than 1%), making each successful case a valuable window into otherwise opaque agent behaviour. We classify 1,254 performance-relevant diff hunks from 216 of these PRs, spanning five agent systems, against the 18-category syntactic mutation taxonomy of Even-Mendoza et al. (2025) using an LLM-as-a-judge pipeline. Three categories dominate: name modification (36.9%), object creation (26.3%), and type change (22.6%), a profile strikingly different from prior genetic improvement corpora where no change accounted for 84%. Each agent commits to a distinctive mutation vocabulary, and each performance strategy activates a largely disjoint category subset. Agent identity and target strategy are therefore informative priors that narrow the effective SBSE operator space from 18 categories to a handful per context. Replication package: https://anonymous.4open.science/r/ssbse-challenge-2026-710C/ 2026-07-10T00:00:00Z http://hdl.handle.net/1893/38347 Title: Bridging Learning Outcomes and Module Specifications: GenAI Framework for Curriculum Development in CS Education Author(s): Elawady, Mohamed; Ali, Hazrat Abstract: Recent developments in generative artificial intelligence (GenAI) have enabled the production of high-quality content in many domains , including education and computing. These advances offer opportunities to support tutors in curriculum development and to enhance the learning experience for students. This paper presents an end-to-end pipeline that automatically suggests a set of module specifications based on user input and structured guidelines. 2026-09-03T00:00:00Z http://hdl.handle.net/1893/38262 Title: Optimizing Appliance Scheduling for Solar Energy Management Us-ing Metaheuristic Algorithms Author(s): Ahmed, Hiba; Brownlee, Alexander E I; Adair, Jason; Powers, Simon T Abstract: Solar energy generation is often misaligned with when households use power, creating a scheduling challenge. We optimize appliance start times in an island microgrid setting to minimize user dissatisfaction while promoting solar use and respecting system constraints. A sequential multi-day scheduling framework using Iterated Local Search (ILS) and Simulated Annealing (SA) considers power consumption, active duration, inverter size, battery limits, and solar forecasts, opening potential to explore trade-offs between cost, system size, and satisfaction. 2026-08-13T00:00:00Z http://hdl.handle.net/1893/38059 Title: Empowering Stakeholders with Participatory Auditing of Predictive AI: Perspectives from End-Users and Decision Subjects without AI Expertise Author(s): Di Campli San Vito, Patrizia; Fringi, Eva; Johnston, Penny; Bezerra, Leonardo C T; Aristodemou, Marios; Shahandashti, Siamak F; O'Hara, Emily; Fiona Whyte, Laura; Luo, Lin; Wong, Mark; Soufan, Ayah; Moshfeghi, Yashar; Stumpf, Simone Abstract: Artificial intelligence (AI) applications have become ubiquitous in their impact on individuals and society, highlighting a crucial need for their responsible development. Recent research has called for participatory AI auditing, empowering individuals without AI expertise to audit AI applications throughout the entire AI development pipeline. Our work focuses on investigating how to support these kinds of auditors through participatory AI auditing tools and processes. We conducted a series of co-design workshops, using two health-related predictive AI applications as examples. Our results show that participants wanted to be part of AI audits, and were insightful in identifying the potential impacts of applications, but needed to be assisted in conducting audits, especially how to measure impacts. Importantly, participants provided examples of impacts not considered in current risk/harm taxonomies. Our findings provide implications for the design of tools and processes to empower everyone to contribute to responsible AI development in the future. 2026-04-01T00:00:00Z