The Summer School on Machine Learning at Skoltech (SMILES-2024), focused on generative AI, has commenced at the historic Tsarskoe Selo venue, now part of the "Kochubey Center" in Pushkin, St. Petersburg. This 12-day intensive course is designed for young professionals, offering them the opportunity to study the latest advancements in machine learning and apply them to solve complex industrial challenges. The school’s principal partner is the Sberbank Laboratory of Artificial Intelligence and Robotics Center, with Alfa-Bank serving as the gold partner. The National Project "Digital Economy" also supports the school.
More than 700 undergraduates, master's students, and Ph.D. candidates from universities across Russia, who are either actively engaged in machine learning or planning to pursue a career in AI, participated in the competitive selection process. Seventy-five students were invited to attend in person, over 200 were invited to participate online, and nearly 1,200 individuals registered to watch the livestreams. Participation in SMILES-2024 is free for all attendees, with accommodation, meals, and in certain cases, travel expenses covered for those attending in person.
The curriculum covers the most advanced tools in machine learning, including large language models, multimodal approaches, multi-agent systems, generative techniques, and self-learning methods. Among the distinguished lecturers are the school's scientific director Evgeny Burnaev (listed among the top 2% most cited scientists in the world by Elsevier in 2023), the Director of the Skoltech AI Center and head of the AIRI research group; Alexander Kraynov, Director of AI Technology Development at Yandex; Vladimir Kramnik, the 14th World Chess Champion; Alexander Gasnikov, Rector of Innopolis University and Chief Research Scientist at the Skoltech Applied AI Center; and Alexander Korotin, Head of the Research Group at the Skoltech Applied AI Center and Research Scientist at AIRI.
Practical workshops will demonstrate how the machine learning tools presented in the lectures play a crucial role in solving real-world industrial problems through engaging scientific challenges. Participants will have the chance to implement the methods and algorithms discussed in their own projects during the "AI Methods for Solving Industrial Problems" hackathon, forming teams of 2–4 people. Today marks the beginning of the project work: 23 in-person teams, 31 online teams, and a few mixed groups. Participants will have access to Skoltech’s supercomputer "Zhores" to aid in their project development.
"For five years, our SMILES school has been attracting top lecturers and students in the field," said Alexey Zaytsev, Director of SMILES-2024, Senior Lecturer at Skoltech, and Head of the LARSS Laboratory at the Skoltech Applied AI Center. "In addition to lectures, we offer cases to solve during the summer school. This year, participants will tackle tasks such as detecting hallucinations in large language models, identifying cheating in chess, and enhancing neural network models for analyzing bank customer behavior. Besides academics, leading industrial companies are actively involved in proposing these challenges. We have ensured that participants gain new knowledge that they can apply in their academic and industrial careers paths."