Meet the heads of research behind Switzerland's fully open-source Apertus LLM for the latest insights from our flagship project: training transparent, compliant, and massively multilingual AI models at scale.

We will share candid experiences from developing the latest 1.5 model, featuring an on-stage live demo that highlights the studies currently paving the way for Apertus 2.0. In a fireside chat, we will unpack the complex trade-offs involved in dataset curation, incorporating community feedback, and maintaining scientific fidelity at scale. You will leave the session equipped with ready-to-run instructions, tutorials, benchmarks, and concrete strategies for weaving Apertus into your own AI roadmap.