DeepMind’s latest research at ICLR 2023

Hey tech enthusiasts! Let’s divdive into some fascinating insightful content for our brand new blog post about recent technical research and breakthroughs. Our latest research has shown that AI can now generalise to new tasks, making it suitable for scientific advancements and complex drug discovery. We’ve introduced several transformer-based models that can accurately predict molecular properties from their 3D structures, and we have also drawn inspiration from physics to simulate collision simulations for robotics, graphics, and mechanical design. Our research teams are presenting a diverse range of papers at this year’s International Conference on Learning Representations (ICLR), including collaborative works with other researchers. As sponsors and regular workshop organizers, our team is dedicated to supporting the event as it takes place in Kihira, Rwanda. We’ve also presented 23 papers, including several collaboration projects this year. Our latest breakthroughs in AI have opened up new avenues for scientific advancements and complex drug discovery. As researchers, we believe that AI can now generalise to new tasks and make significant contributions in these fields. Our work is essential as they are fundamental to our understanding of the world around us. We encourage you to explore our blog post further and stay tuned for more updates on our latest research!