NeoMundi Research studies how runtime signals can make generative AI behavior observable, comparable and governable across models, tasks and professional contexts. NeoMundi doesn't just publish scores. NeoMundi observes how a runtime signal becomes interpretable, open to criticism, and useful for the AI ecosystem.