Leiden, The Netherlands – We are thrilled to announce that we have two papers nominated for the Best Paper Award at GECCO 2026, the premier venue in evolutionary computation.
🏆 Best Paper Award Nominees
Assessing Reproducibility in Evolutionary Computation
Authors: Francesca Da Ros, Tarik Začiragić, Aske Plaat, Thomas Bäck, Niki van Stein
How reproducible is EC research, really? We studied 10 years of GECCO papers, built a structured reproducibility checklist, and developed RECAP: an LLM pipeline that automatically audits reproducibility of published work. Our findings reveal clear room for improvement in the field — but also show that automation can help address this challenge at scale.
Structural Bias in Continuous Multi-objective Optimization
Authors: Jakub Kůdela, Niki van Stein, Thomas Bäck, Anna Kononova
Optimization algorithms can be systematically biased towards certain regions of the search space — regardless of the objective function. We extend this concept of structural bias to the multi-objective setting for the first time, introducing a suite of synthetic test problems designed to isolate purely algorithmic bias.
These results highlight our group’s focus on two intertwined themes: LLMs as engines for automated algorithm design, and rigorous evaluation and benchmarking of evolutionary computation methods. We look forward to presenting this work at GECCO 2026 in San José, Costa Rica 🇨🇷
Follow us on LinkedIn for updates, preprints, and post-conference impressions.