This track explores the challenges and opportunities in designing software system infrastructure for training and serving large-scale AI models. It covers key aspects such as scalability, resource efficiency, system optimization, and deployment strategies, aiming to foster discussions on practical solutions, emerging techniques, and best practices essential for efficient and efficient AI infrastructure.

AI + Education is a multi-layered topic. In this session, we will explore the various ways AI is influencing and transforming both how we educate and what we teach.

We will begin with an overview of the different dimensions of this complex field and look at it from different perspectives:

- The research perspective – focusing on the foundational educational research surrounding the use of AI in schools and universities.

- The development perspective– focusing on the development of AI tools for education.

- The ETH Zurich perspective– presenting how the university is building Ethel, an open-source platform designed to support both lecturers and students.

- The AI Challenge – a project that encourages teenagers in exploring AI through a creative, hands-on project.

The session will conclude with a panel discussion focusing on the role of the ETH AI Center within this dynamic and evolving landscape and how it can contribute to an AI ready education in Switzerland.

This track explores AI's impact on science and engineering, featuring presentations on various topics ranging from chemistry, fluid dynamics, engineering to neuroscience. A panel will discuss AI's capabilities and challenges, ending with insights on using AI to reduce engineering costs.

Artificial intelligence is revolutionizing Industry 4.0 by transforming static automation into dynamic, self-optimizing systems that learn and adapt in real-time. This track introduces AI-based methods for automation in manufacturing, demonstrating how intelligent optimization and adaptive decision-making enhance traditional systems. Key directions include AI-enhanced process simulations, optimization, and control algorithms that enable manufacturing systems to learn from data, respond to changing conditions, and continually improve performance. Presentations will feature novel approaches, real-world case studies, and strategic insights for integrating AI into existing industrial environments, helping attendees understand how to select suitable AI methods and transition from conventional automation to intelligent manufacturing.

Computer Vision aims to endow AI agents with human-level understanding of human performance. To build agents with such advanced machine perception, we must build computational methods and representations to model how humans move within the physical world and interact with it. This track aims to give an overview of efforts in this domain, focusing on different aspects of the machine perception problem.

The rise of large language models (LLMs), pretrained on vast and diverse datasets, has revolutionized the field of artificial intelligence (AI). Finetuning has emerged as a critical next step for adapting these models to a wide range of downstream applications, serving as the “last mile” for various use cases. Compared to pretraining LLMs from scratch, finetuning open-source models offers a more accessible and practical alternative for small to medium-sized businesses and academic researchers, who might not have access to extensive computational resources. Despite its promise, the broad applicability of finetuning also introduces several challenges.

This workshop focuses on efficiency in fine-tuning LLMs, aiming to lower costs and barriers so that even users with consumer-grade GPUs can harness reasonably large models (e.g., 7B parameters). Our goals are twofold: (i) to enable scalable development of LLMs, and (ii) to empower individuals and organizations with limited resources to benefit from modern AI. We will convene recent advances in methods and tools, and foster the exchange of best practices across research and industry.