We are pleased to announce the accepted EvoApplications special session “Beyond a Performance Scalar: Evaluation Science for Evolutionary Computation".

The session will explore how evolutionary algorithms can be evaluated beyond a single fitness value or ranking. It will focus on richer evidence about algorithm behaviour, including search dynamics, statistical significance, robustness, uncertainty, generalisation, computational cost, reliability, and practical relevance.

Contributions are invited on topics such as anytime and time-to-target evaluation, convergence and trajectory analysis, diversity and population dynamics, mechanism attribution, ranking stability, multi-criteria evaluation, and evaluation in noisy, dynamic, constrained, or resource-limited settings. Relevant work in evolutionary machine learning, AutoML, automated algorithm design, and LLM-assisted evolutionary computation is also welcome.

The session is organised by Rohit Salgotra, Tome Eftimov, Niki van Stein, Anja Janković, and Eva Tuba.

Researchers interested in making evolutionary computation more rigorously evaluated, interpretable, and practically meaningful are warmly invited to contribute and join the discussion at EvoApplications.