Lukas Helff

As a research scientist at the Hessian AI Service center, my work spans across the domains of computer vision, natural language processing, and multimodal AI systems, with a primary focus on enhancing safety within artificial intelligence. We develop and deploy input-output safeguard models which do help us to enhance the safety and integrity within machine learning datasets as well as Human-AI interactions by effectively detecting and mitigating unsafe or inappropriate content. Further research involves evaluating and enhancing visual and logical reasoning capabilities of AI systems. To this end, we employ and refine a broad spectrum of AI methodologies, including neural AI, symbolic AI, and Neural-Symbolic AI, aiming to create more robust and reliable AI technologies that can positively impact various domains.