11.11.2026
In early November 2026, the BilAI Cluster of Excellence opened its third retreat at TU Graz. 100 researchers from all over Austria gathered to advance collaborative research on Bilateral AI through scientific exchange, discussion and networking.
The two-day retreat set the course for the Cluster’s next research phase, guided by an updated research strategy. To keep the Cluster well positioned in a rapidly evolving research environment and to strengthen cooperation, a substantial amount of the cluster’s work is bundled in three “light towers”.
All three focus on the integration of symbolic and sub-symbolic AI. The BilAI Agentic Systems light tower is developing robust, rule-abiding and safer AI agents based on symbolic representations such as knowledge graphs. BilAI LLM develops efficient new model architectures and symbolic extensions for foundation models. Lastly, the BilAI Optimizer, an example of a game-changer application that equips agents and language models with exact planning, reasoning and optimization capabilities, while neural models help improve the underlying algorithms. During the retreat and beyond, each light tower serves as a meeting point for the research community.
Interactive discussions in focus groups during workshop sessions laid the groundwork for new collaborations. These in-person discussions resulted in the formation of new research teams, with the aim of addressing open questions and advancing interdisciplinary work in BilAI. Next to lightning talks showcasing bilateral work from the community, 15 doctoral researchers presented their ongoing projects and results, highlighting their progress. Wherever their projects could benefit from joint efforts, the presenters invited their fellow BilAI researchers to connect across fields and methodologies, leading to lively discussions in the poster sessions and networking breaks.
Together with self-organized mini-workshops and discussion sessions held the day after the retreat, this made for a highly productive gathering of the Cluster’s community. Both established and new collaborations were fostered and many new ideas and connections emerged across BilAI.