[EN] AI systems are usually evaluated with a variety of benchmarks to determine their performance for specific tasks, using a single metric which provides a simplistic image of their capabilities. However, this procedure ...